Audience Activity and

Transcription

1 Audience Activity and Soap Opera Involvement A Uses and Effects Investigation ALAN M. RUBIN Kent State Uniuersity ELIZABETH M. PERSE University of Delaware This investigation examined the roleofmotives, attitudes, andaudienceactivity in explaining the affective, cognitive, and behavioral involvement of 328 daytime soap opera viewers. Because intercorrelations were found among motives, attitudes, activities, and involvement variables, canonical correlation analysis was used. There were two multivariatepattems. First, exceptfor viewing topass time, more salient viewing motivations (especially exciting entertainment and social utility), perceived realism, viewing intention, and attention were related to parasocial interaction, postviewing cognition, and postviewing discussion. Second, viewingfor social utility, but notfor voyeurism, and the lack of realism were related to postviewing discussion, but not to parasocial interaction. These audience orientations and the role of involvement in media uses and effects were discussed. U ses and gratifications have moved into a major phase of development marked by vigorous theoretical growth and research that fills earlier conceptual and empirical gaps (Palmgreen, 1984). Nowhere have such gaps been more apparent than the separate treatments of gratifications and effects research. In 1981, Windahl argued for a merger of these two research traditions. The present study integrated uses and effects perspectives by examining audience involvement with media content. Alan M. Rubin (Ph.D., University of Illinois at Urbana-Champaign, 1976) is associate professor in the School of Speech Communication and the School of Journalism and Mass Communication at Kent State University. Elizabeth M. Perse (Ph.D., Kent State University, 1987) is assistant professor in the Department of Communication at the University of Delaware. An earlier version of this article was presented at the Speech Communication Association convention, Chicago, November Human Communication Research, Vol. 14 No. 2, Winter o 1988 International Communication Association 246

2 Rubin, Perse / SOAP OPERA INVOLVEMENT 247 A primary tenet of uses and gratifications is that audience members shape their own media experiences. Blumler (1979) described audience activity as an important intervening variable in media effects. Audience activity connotes goal-directed intentional and selective media use. By understanding the variable nature of audience activity in the mass communication process we should be better able to explain media effects. In the present investigation we attempted to move beyond research that has identified types of activity and their relationships to media exposure patterns (e.g., Levy, 1978, 1983; Levy & Windahl, 1984). To explicate the role of activity in media uses and effects, this study focused on college students involvement with daytime television soap opera content. We examined the role of motives, attitudes, and activities in affective, cognitive, and behavioral involvement with that content. Involvement Znuoluement has two meanings in communication research. One view is derived from persuasion and marketing research. Growing out of social judgment theory (Sherif, Sherif, & Nebergall, lw), involvement is seen as a precommunication condition that reflects what people bring with them to the communication situation. It is a motivational state linked to strength and centrality of preexisting attitudes. Involvement reflects perceived importance of information; it influences people s reactions to messages (Sherif et al., 1965). When faced with messages about important topics, people are motivated to pay attention to informational message elements and process them more intensely (Petty & Cacioppo, 1984). Krugman (1966, p. 583), though, pointed to a second meaning of involvement: direct personal experience during message reception. As reflected in interpersonal, political, and mass communication research, involvement indicates participation. When people are involved they focus on the message (Bailyn, 1959) and react emotionally (Capella, 1983; Miyazaki, 1981). Involved television viewers pay attention to and think about the message, get caught up in the action of the drama (Bryant & Comisky, 1978, p. 65), and identify with television characters (Rosengren & Windahl, 1972). Involvement, then, is cognitive, affective, and behavioral participation during, and because of, media exposure. Participant involvement includes parasocial interaction with television personalities, and thinking about and discussing media messages.

3 248 HUMAN COMMUNICATION RESEARCH / Winter 1987 Affective involvement. Parasocial interaction is affective participant involvement. Conceptualized as a sense of friendship formed by audience members with media personalities, parasocial interaction is based on the frequent and consistent appearance of a realistic persona (Horton & Wohl, 1956). Encouraged by conversational manner, interpersonal style, and media production techniques, viewers may react interpersonally to television personae and feel they know the characters the way they know their friends. McGuire s (1974) discussion of theories that undergird uses and gratifications suggests that parasocial interaction may be explained by active, affective theories of affiliation that portray people as being basically altruistic and cohesive (p. 188) and motivated to form bonds with others. Parasocial interaction is a sense of involvement with media personalities based on normal interpersonal expectations and processes (Rubin & McHugh, 1987; Rubin, Perse, & Powell, 1985). Cognitive involvement. Thinking about messages is cognitive participant involvement. The cognitive response approach to persuasion has demonstrated that people think more about important than unimportant messages (Greenwald, 1968; Petty & Cacioppo, 1979). Greenwald and Leavitt (1984) suggested that thinking about messages reflects elaboration, a deeper level of involvement. This view is echoed by Levy and Windahl(l985) and Rosengren and Windahl(l972) who proposed that involvement is marked by thinking about media messages during and after exposure. Behavioral involvemen t. Talking about messages is behavioral participant involvement. For example, Tan (1980) found that interpersonal discussion led to the use of newspapers during a political campaign. Stanford (1984) observed that those who had a personal guidance orientation to television were the most involved, discussing characters and plots and talking back to the television. Bechtel, Achelpohl, and Akers (1972) and Lemish (1985) observed that the most involved viewers were those who spoke with others about program content. Audience Activity and Involvement Audience activity is multidimensional (Blumler, 1979). According to Levy and Windahl (1984, 1985), audiences exhibit different levels of activity before, during, and after exposure. Researchers have found associations among types of activity and viewing motives and behaviors.

4 Rubin, Perse / SOAP OPERA INVOLVEMENT 249 For example, Rubin and Perse (1987) found two multivariate local news viewing orientations. Instrumental use focused on news content and was marked by information- and exciting entertainment-seeking viewing motives, greater perceived realism and news affinity, intentional viewing, and cognitive involvement with the news. Ritualistic use focused less on the news content than on exposure to the medium; it was marked by watching out of habit or to fill time, lower levels of news affinity, selectivity, and intentionality, but higher levels of coviewing distraction. As types of audience activity, viewing intention and attention have been linked to cognition and behavior. For example, goal-directed viewers check the time and arrange their schedules to watch a program (Lemish, 1985; Levy 8z Windahl, 1985). More active viewers attend to the personal qualities of television characters, and care and talk about the characters (Rouner, 1984). As a motivated state (Kellermann, 1985), attention to plot and characters indicates greater cognitive activity when viewing a program and deeper processing of program content (Levy & Windahl, 1985). As components of an instrumental media use orientation, greater viewing intention and attention should be linked to enhanced opportunities for affective, cognitive, and behavioral involvement. Gratifications, Activity, and Effects Levy and Windahl (1984, pp ) speculated that more active individuals not only receive higher amounts of gratification from their media use, but also that they are more affected by such active and gratifying exposure. Studies that considered links among viewing patterns and outcomes observed instrumental television use to be associated with soap opera involvement (Rubin, 1985), limited soap opera cultivation effects (Perse, 1986), and parasocial interaction with local news personalities (Rubin et al., 1985). Although Levy and Windahl (1985) speculated that parasocial interaction reflects reduced audience involvement, it is likely that parasocial interaction indicates active and involving television use. Wenner (1983), for example, reported that parasocial interaction with network newscasters was related to an affective search for vote guidance, interest in political campaigns, and political discussion. Rubin and his associates (1985) found information-seeking viewing motives to predict parasocial interaction with local television newscasters.

5 250 HUMAN COMMUNICATION RESEARCH / Winter 1987 Other researchers have observed cognitive and behavioral outcomes of active media use. For example, Gantz (1978) found news attentiveness and information viewing motives to contribute to television news recall. A need for political orientation (Weaver, 1977) and attentive news viewing (Hill, 1985) have been linked to agenda setting. McLeod and Becker (1974) found that increased campaign activity, political interest, and voting intention were associated with watching political programming for surveillance, reinforcement, vote guidance, and communication utility. Media effects also have been linked to attitudes, especially how important and realistic television is. Perceived realism, for example, is an important ingredient in cultivation effects (Perse, 1986; Potter, 1986). And, perceived realism and news affinity predict parasocial interaction with television newscasters (Rubin et al., 1985). Based on past research, then, we expected instrumental viewing motives, perceptions of affinity and realism, and two types of audience activity-viewing intention and attention-to be linked with affective, cognitive, and behavioral involvement. As compared to less active consumers, instrumental viewers should be affected more by some television content because they are motivated to select that content, perceive it to be realistic and important, pay attention to it, and integrate it into their conversations and behavioral intentions. Although research has focused on information content, few uses and effects studies have examined entertainment genres. Soap Operas and Involvement Soap opera audiences exhibit viewing behaviors that reflect heightened activity and involvement. Soap operas have the highest rate of repeat viewing (Barwise, Ehrenberg, & Goodhardt, 1982), and selective audiences who choose programs instead of stations (Cantor & Pingree, 1983). Viewers talk to characters on the screen or to each other about the content (Lemish, 1985; Whetmore & Kielwasser, 1983). Attention levels when watching daytime serials are higher than for other daytime (McDonald, 1974) or prime-time programs (Cantor & Pingree, 1983). Whetmore and Kielwasser (1983) reported soap opera viewer involvement on several levels: looking for clues to predict plot outcomes, attention to fashion, identification with a favorite character, and observation of interpersonal dynamics. Production techniques, dialogue, music, lighting, and camera move-

6 Rubin, Perse / SOAP OPERA INVOLVEMENT 251 ments are designed to create involvement (Barbatsis, 1983; Cassata, 1983; Timberg, 1983). Content and production techniques encourage viewers to relate to characters, to think about stories, and to discuss content. Plot complexity, crises, and an emphasis on conversation encourage cognitive and behavioral involvement, especially thinking about and discussing stories and trying to predict future outcomes (Lemish, 1985; Whetmore & Kielwasser, 1983). Soap operas rely on conversation to advance plots and provide viewers with secondhand self-disclosure and an opportunity to get to know the characters (Fine, 1981; Katzman, 1972; Turow, 1974). The basic staple of soap operas, the development of personal problems encountered by attractive characters, encourages affective involvement. Audience members are invited to participate in the experiences of characters through several mechanisms: the central role of characters in plots, the insight given into how characters think and feel, the resemblance of characters to everyday people, and the time spent on character history and plot development (Arnheim, 1944; Rose, 1979). Techniques of introducing soap opera involvement can be effective with a variety of audiences. Recent studies have found college students to be a loyal and active soap opera audience (Carveth& Alexander, 1985; Perse, 1986; Rubin, 1985). Similar to other audiences, college students watch soap operas for reality exploration, avoidance, diversion, and social utility (Rubin, 1985). Many are intentional and attentive soap opera viewers, even arranging class schedules or seeking televisions on campus so as not to miss their favorite stories (Lemish, 1985). Building upon earlier studies,? then, a college student sample provides an appropriate focus for examining audience activity and soap opera involvement. Research Model The present study was designed to examine an instrumental media uses and effects model: involvement results from media use motives filtered by individual attitudes and audience activity (Rubin & Perse, 1987). Greater affective, cognitive, and behavioral involvement with soap opera content should be linked to an instrumental orientation toward daytime television serials. These three manifestations of involvement were operationalized, respectively, as parasocial interaction with a favorite soap opera character, thinking about the content

7 252 HUMAN COMMUNICATION RESEARCH / Winter 1987 of a favorite soap opera, and discussing the content of a favorite soap opera. We expected that parasocial interaction, postviewing cognition, and postviewing discussion would be related positively to (a) instrumental soap opera viewing motives, such as social utility and seeking exciting information, (b) soap opera attitudes or perceptions of soap opera affinity and realism, and (c) audience activity, including intention to watch a favorite soap opera and attention to the program content when viewing. Procedure and Sample METHOD Building on a systematic line of research, we focused on the use of daytime television soap operas by college students. Similar to past studies (e.g., Carveth & Alexander, 1985; Perse, 1986), questionnaires were administered during class periods to 460 undergraduate students at a large midwestern university in fall Of the total sample, 71.3% said they watched daytime television soap operas. This analysis was based on that sample of 328 soap opera viewers. Those in this sample ranged in age from 17 to 34 (M= years, SD = 2.09); 8.4% were freshmen, 32.2% sophomores, 37.2% juniors, and 22.3% seniors. The sample was 68.9% female. Most (96.6%) were never married. On the average, respondents watched 2.76 (SD = 1.91) hours of television daily, viewed their favorite daytime soap opera 2.60 (SD 1.41) times per week, and had been watching this program for4.79 (SD = 3.14) years. The most favored serials were GeneralHospital(38.6% of responses), All My Children (19.6%), Days of Our Lives (13.1%), and the Guiding Light (10.6%). Measurement Viewing motives. Respondents were asked about their agreement with 30 reasons for watching their favorite daytime soap operas. Five response options, ranging from not at all (1) to exactly (5) like their own reasons for viewing, were used. The statements were adapted from prior research (e.g., Rubin, 1981,1983; Rubin et al., 1985). Attitudes. Adaptingitems shown to be reliable in earlier studies (e.g., Rubin, 1981,1983; Rubin et al., 1985), two scales were used to gauge soap opera attitudes. First, a 5 item affinity scale measured the perceived

8 Rubin, Perse / SOAP OPERA INVOLVEMENT 253 importance of watching favorite daytime television serials. For example, I would rather watch my favorite daytime soap opera than watch anything else on TV, Whenever I m unable to watch my favorite soap opera, I really miss it, and Watching my favorite soap opera is one of the more important things I do each day. Second, a 6-item realism scale measured how true-to-life respondents perceived their favorite daytime television serials to be. For example, My favorite soap opera lets me see how other people live, My favorite soap opera helps me understand some of the problems other people have, and My favorite soap opera presents things as they really are in life. For each scale, respondents indicated their agreement with statements across five response options ranging from strongly disagree (1) to strongly agree (5). Audience activity. Three measures were employed to assess cognitive and behavioral audience activity before and when watching a favorite daytime television soap opera. These measures were inten tionality or planning to watch one s favorite soap opera, attention to the program when watching, and engaging in distracting behaviors when watching. A 4-item intentionality scale, expanded from Levy and Windahl s (1984) preexposure activity measures, concerned a person s behavioral intention to watch a favorite serial. Items reflected the following: felt importance of watching the program from beginning to end, planning days so as not to miss a favorite soap opera, checking times so as not to miss the program, and planning to see the program from start to finish. Response options ranged from strongly disagree (1) to strongly agree (5). Rouner s (1984) four items were used to assess viewing attention to the content and characters of one s favorite soap opera. Respondents were asked to indicate the amount of attention they usually gave to the story line, the personal qualities of the characters, the appearance of characters, and the values and morals displayed by the characters. These bipolar measures ranged from no attention at all (coded 1) to very close attention (coded 5). In contrast to the attention scale, engaging in couiewing distractions reflects reduced program attention and involvement (Levy, 1978,1983). Based on several Levy and Windahl (1984) dur-activity items, these eight measures assessed behavior that would distract a viewer from program content. Respondents were asked how often they engaged in several distractions while watching their favorite soap opera: reading

9 254 HUMAN COMMUNICATION RESEARCH / Winter 1987 newspapers, books, or magazines; preparing food; eating; doing housework or chores; doing needlework or hobbies; talking about things other than what s on the soap opera; doing homework or paperwork; and daydreaming or thinking about other things. Response options ranged from never (1) to very often (5). Involvement. The conceptual model of the investigation proposed affective, cognitive, and behavioral involvement with soap operas. These three dependent measures, expected to covary with different viewing motives, attitudes, and activities, were parasocial interaction, postviewing cognition, and postviewing discussion. Parasocial interaction was conceptualized as a sense of affective interpersonal involvement with media personalities. The Parasocial Interaction Scale was adapted for television soap operas from a previous investigation of local television news viewing (Rubin et al., 1985). Statements reflecting perceptions of and dispositions toward a favorite television soap opera character were presented to the respondents who indicated their agreement according to five response options, which ranged from strongly disagree (1) to strongly agree (5). Item analysis indicated that the original 20-item scale could effectively be reduced, producing a 10-item Revised Parasocial Interaction Scale (PSI).* To consider an individual s cognitive involvement in processing program content, respondents were asked to indicate their agreement with four items about the story and characters after they watched their favorite soap opera. The postviewing cognition scale reflected the following: thinking about what happened in the story, thinking about what was seen or heard, trying to predict what will happen in tomorrow s episode, and thinking about the characters. Response options ranged from strongly disagree (1) to strongly agree (5). To assess a person s behavioral involvement in sharing program content, respondents were asked to indicate their agreement with three statements about the story and characters after they watched their favorite soap opera. The postuiewing discussion scale reflected the following: talking about the story with others, talking about the characters with others, and talking to others to try to predict what will happen in the story. Response options ranged from strongly disagree (1) to strongly agree (5). Statistical Analysis Scale construction. Because media use motives are interrelated (Rubin, 1983,1984; Rubin & Perse, 1987; Rubin et al., 1985), soap opera

10 Rubin, Perse / SOAP OPERA INVOLVEMENT 255 viewing motives were determined from principal axis factoring (PAF) with iterations and oblique rotation (SPSS, 1986). A factor required a minimum eigenvalue of 1.0 and at least three primary loadings of.40 or greater to be retained. Factor scores were computed via the regression method and used to create the six soap opera viewing motives. Oblique-rotated PAF analysis (SPSS, 1986) of the 10 parasocial interaction items supported a single-factor solution accounting for 51.0% of the total variance. Responses to the 10 PSI items, therefore, were summed and averaged. In addition, responses to the items of each remaining construct were summed and averaged to form the respective attitude, activity, and involvement scales. Conceptual tests. Following from factor and reliability analyses to construct the several scales, two statistical techniques were applied to the data. First, third-order partial correlations, controlling for age, gender, and amount of television viewing, were computed to consider relationships among the variables. Second, given the interrelations among predictor and criterion variables, canonical correlation analysis was used to assess the multivariate relationships among the set of antecedent variables (viewing motives, attitudes, and activities) and the set of consequent involvement variables (parasocial interaction, postviewing cognition, and postviewing discussion). The canonical analysis included computation of canonical loadings (i.e., structure coefficients) and redundancy coefficients. Because canonical loadings are largely free from the direct influence of multicollinearity (Lambert & Durand, 1975, p. 472), they remedy possible misinterpretation caused by multicollinearity affecting canonical weights (Levine, 1977; Tucker & Chase, 1980). Redundancy coefficients provide the proportion of variance of one set accounted for or shared by the canonical variate of the other set (Levine, 1977). In other words, redundancy reflects the proportion of variance in one [set] that can be predicted by or attributed to variation in the other (Lambert & Durand, 1975, p. 469). Measurement RESULTS Viewing motives. The PAF solution yielded six interpretable factors accounting for 55.3% of the total variance. A seventh factor had an eigenvalue of only.64. Signs on Factors 3,4,5, and 6 were reversed to reflect the underlying voyeurism, escapist relaxation, information, and

11 256 HUMAN COMMUNICATION RESEARCH / Winter 1987 social utility dimensions. Table 1 summarizes the primary factor loadings. Factor 1, exciting entertainment, had an eigenvalue of 7.16 and explained 11.3% of the total variance after rotation. Three entertainment, one habit, and two arousal items loaded on the factor, which depicted watching soap operas for enjoyment and excitement. One habit and three pass-time items loaded on Factor 2, pass time, which had an eigenvalue of 3.11 and explained 6.2% of the total variance. The factor denoted ritualistic soap opera viewing to occupy one s time when there was nothing better to do. Factor 3, voyeurism, had an eigenvalue of 2.14 and explained 6.9% of the total variance. It included three items reflecting the sexual appeal and attractiveness of the characters and program. Factor 4, escapist relaxation, had an eigenvalue of 1.40 and explained 9.0% of the total variance. It contained three relaxation and two escape viewing items. The factor depicted soap opera viewing to unwind, relax, and forget about work and other things. The three information items loaded on Factor 5, information, which had an eigenvalue of 1.11 and explained 4.8% of the total variance. It reflected information seeking to learn things about oneself or others. With secondary loadings on the voyeurism factor, this was a marginal factor. The three social interaction items loaded on Factor 6, social utility, which had an eigenvalue of 1.04 and explained 5.1% of the total variance. It indicated a more instrumental use of soap operas to be with and interact with others. The mean scores of the primary loadings of the six factors indicated that viewing for exciting entertainment (M= 3.61, SD =.80, Cronbach alpha =.87), for escapist relaxation (M=3.16, SD=.87, Cronbach alpha =.84), and to pass the time (M= 2.77, SD =.89, Cronbach alpha =.73) were the primary viewing motives. Viewing for social utility (M = 2.25, SD =.89, Cronbach alpha =.71), for voyeurism (M = 2.03, SD = 87, Cronbach alpha =.84), and for information (M = 1.60, SD =.70, Cronbach alpha =.78) were less dominant reasons for watching soap operas. Attitudes. The 5-item affinity scale (M = 1.92, SD =.73) had a.80 Cronbach alpha. The 6-item realism scale (M= 1.86, SD=.69) had a.87 Cronbach alpha. The mean scores indicated that favorite soap operas were not perceived to be particularly realistic or important. Audience activity. The 4-item intentionality scale (M= 2.29, SD =.97) had a.87 Cronbach alpha. The 4-item attention scale (M = 3.77, SD =

12 Rubin, Perse / SOAP OPERA INVOLVEMENT ) had a.76 Cronbach alpha. The 8-item coviewing distraction scale (M = 3.19, SD.69) had a.76 Cronbach alpha. Involvement. The 10-item Parasocial Interaction Scale (M 2.67, SD =.80) had a.88 Cronbach alpha. Scores on the 10-item PSI correlated strongly with scores on the original 20-item scale (r =.96,p <.001). The 4-item postviewing cognition scale (M = 2.74, SD =.86) had a.86 Cronbach alpha. The 3-item postviewing discussion scale (M= 2.97, SD = 1.01) had a.89 Cronbach alpha. Correlates of Involvement The first stage of the analysis was to examine the correlations among the variables, controlling for age, gender, and amount of television viewing. Given the anticipated interrelated nature of the constructs of the investigation, the many modest to moderate partial correlations were not surprising. The partial correlation matrix is reported in Table 2. The involvement outcomes had several sizable correlates. The strongest of these associations was between postviewing cognition and discussion; PSI was related more modestly to the two other involvement variables. PSI, cognition, and discussion also were related moderately to the exciting entertainment viewing motivation, soap opera affinity, and viewing intention. PSI and cognition also exhibited moderate associations with perceived realism and viewing attention. There also were moderate relationships among other variables: the exciting entertainment motive with viewing intention, affinity, viewing attention, and the escapist relaxation motive; the information motive with perceived realism; affinity with viewing intention, perceived realism, and viewing attention; perceived realism with viewing attention and viewing intention; and viewing intention with viewing attention. Of particular interest was the distinct and less goal-directed pattern of the ritualistic pass-time motive in comparison to all other soap opera viewing motives. Consistent with the earlier conceptual discussion, then, the three involvement variables were moderately intercorrelated (interitem Pearson r = S6). Bartlett s test of sphericity reinforced the significant association among the three involvement measures: x* (3, N = 320) = , p <.001. These tests supported the appropriateness of the multivariate canonical correlation analysis.

13 TABLE 1 Primary Factor Loadings of Soap Opera Viewing Motives Viewing Motive Items Viewing Motive Fuctors I Watch My Fuvorite Daytime Television Soap Opera.., I ENTR PSTM VOYR RELX INFO SOCL Factor 1 : Exciting Entertainment Because it s enjoyable (3.91, 0.91 ) Because it entertains me (3.94,0.87) Because it s exciting (3.26, 1.16) Because it s thrilling (3.09, 1.1 6) Because I just like to watch (3.75, 1.12) Because it amuses me (3.69,0.94) Factor 2: Pass Time Because it passes the time away, especially when I m bored (3.01, 1.19) When I have nothing better to do (2.92, 1.23) Just because it s on (2.56, 1.28) Because it gives me something to do to occupy my time (2.57,l.lo) Factor 3: Voyeurism Because of the sex appeal in the program (2.21, 1.03) l O oo l O 3.14.O oo.04.oo

14 Because the characters are sexually attractive (2.30, 1.12) Because I find it sexually arousing (1.60,0.84) Factor 4: Escapist Relaxation Because it allows me to unwind (3.36, 1.07) So I can forget about work or other things (3.34,1.16) Because it s a pleasant rest (3.23, 1.1 3) Because it relaxes me (2.99, 1.03) So I can get away from what I m doing (2.90, 1.18) Factor 5: Information So I can learn about what might happen to me (1.47,0.77) So I can learn how to do things I haven t done before (1.43,0.75) Because it helps me learn things about myself and others (1.88,0.98) Factor 6: Social Utility So I can be with other members of the family or friends who are watching (2.14, 1.12) Because it s something to do when friends come over (2.08, 1.12) So I can talk with other people about what s on (2.54, 1.1 3) oo.oo -.09.O NOTE: The numbers in parentheses are each item s mean and standard deviation. ENTR = exciting entertainment; PSTM = pass time; VOY R = voyeurism; RELX = escapist relaxation; INFO = information; SOCL =social utility oo oo o o 1.o SO -.42.oo.11 -.lo.o w u)

15 Exciting Entertainment - Pass Time Voyeurism TABLE 2 Partial Correlations among Motives, Attitudes, Activities, and Involvement Outcomes ENTR PSTM VOYR RELX INFO SOCL AFFN REAL INTN ATTN DSTR PSI COGN Escapist Relaxation Information.2 1.O Social Utility Affinity Perceived Realism Viewing Intention Viewing Attention.47 -.lo Coviewing Distraction Parasocial Interaction Postviewing Cognition Postviewing Discussion.44 -.l NOTE: These data are thirdarder partial correlations controlling for age, gender, and amount of television viewing. r =.ll,p <.05;r =.15,p <.01; r =.19,p <,001 (two-tailed).

16 Rubin, Perse / SOAP OPERA INVOLVEMENT 261 Canonical Analysis of Soap Opera Uses and Involvement Canonical correlation, then, was used to examine the multivariate relationships between the two sets of variables: soap opera use and involvement outcomes. The canonical analysis is summarized in Table 3, which includes the canonical loadings or structure coefficients and redundancy coefficients. Two significant canonical roots (p <.001) were identified, each demonstrating a different instrumental pattern of daytime television soap opera use. Because canonical loadings below about.30 may be unstable and the relationships ascribed to them are sample specific, due to chance, or the result of extraneous factors (Lambert & Durand, 1975, p. 473), interpretation focused on loadings of.30 or greater in absolute value. The first canonical root (R, =.81) explained 64.8% of the common variance between the canonical variates. The canonical loadings revealed that several viewing motives, attitudes, and activities were interrelated in set one. In particular, watching the serials for exciting entertainment, social utility, escapist relaxation, information, and voyeurism, as well as soap opera affinity, perceived realism, viewing intention, and viewing attention correlated positively. And, the involvement variables were intercorrelated in set two. Parasocial interaction, postviewing cognition, and postviewing discussion correlated positively. Across the two sets, then, except for viewing to pass time, more salient viewing motivation, heightened attitudes, and greater activity were linked positively and strongly to greater affective, cognitive, and behavioral involvement with the soap opera characters and stories. The second canonical root (R, = SO) explained 24.5% of the common variance between the canonical variates. The canonical loadings indicated that the first root was dominated by the social utility viewing motive, which was linked negatively to the voyeurism viewing motive, perceived realism, and viewing attention (and to information and escapist relaxation motives to a lesser degree). In the second set, postviewing discussion and parasocial interaction were related negatively. Across the two sets, then, watching soap operas for social utility reasons was linked to discussion of program content, instead of a sense of parasocial interaction. Or, parasocial interaction with soap opera characters was linked to watching for voyeuristic reasons, without seeking social utility gratifications from soap opera viewing. Redundancy coefficients, though, were small.

17 262 HUMAN COMMUNICATION RESEARCH / Winter 1987 TABLE 3 Canonical Analvsis of SoaD ODera Uses and Involvement Motives, Attitudes, and Activities Exciting Entertainment Pass Time Voyeurism Escapist Relaxation Information Social Utility Affinity Perceived Realism Viewing Intention Viewing Attention Coviewing Distraction Canonical Loading Root 1 Root lo Redundancy Coefficients i.22 1 [.021 Involvement Outcomes Parasocial Interaction Postviewing Cognition Postviewing Discussion O Redundancy Coefficients i.441 ~041 NOTE: Root 1: R, =.81, Rc2 =.65, lambda =.26, F(33,861) = 15.33, p <.001. Root 2: Rc = 50, Rcz =.25, lambda =.73, F(20,586) = 5.03, p <.001. DISCUSSION One goal of this study was to examine the contribution of audience activity to affective, cognitive, and behavioral involvement with media content. The results support Blumler s (1979) belief that activity plays an important intervening role in media effects. The intention and attention activity measures were significant contributors in addition to motives and attitudes to explaining involvement. Parasocial interaction, thinking about content, and discussing content represent related, yet different, dimensions of media involvement. For example, viewing attention and perceived realism of soap opera content correlated with parasocial interaction. This suggests that parasocial involvement reflects attention to realistically perceived content during the viewing experience. The links with viewing motives also reflect differences in involvement dimensions. Those who were motivated to seek social utility gratifications discussed the stories and

18 Rubin, Perse / SOAP OPERA INVOLVEMENT 263 characters with others after watching. The sexual appeal or attractiveness of the characters, however, motivated those who parasocially interacted with the characters. Perhaps, on the other hand, the lack of perceived sexual appeal impedes the development of parasocial relationships. The model that guided the study expected involvement to stem from more instrumental and active use of soap opera content (Rubin & Perse, 1987). The analyses supported the expectation. Parasocial interaction and postviewing cognition and discussion were linked to more instrumental viewing motives, stronger attitudes, and greater activity. This corroborates recent reconceptualization of the role of activity in media effects (Levy & Windahl, 1984, Rubin & Perse, 1987). Contrary to earlier beliefs that less active and ritualistic viewers were more affected by television content (Blumler, 1979; Rubin, 1984; Windahl, 1981), more active viewers experience higher levels of gratification. This more active viewing orientation was linked to involvement in this investigation. Rubin and Perse (1987) found that ritualistic and instrumental television news viewing orientations were associated with variations in audience activity. One interesting finding of the present study was that an instrumental orientation to media use can be multidimensional. This supports McLeod and Becker s (1981, p. 72) belief that motives or gratifications sought are only part of what audience members bring to the effects equation. The multivariate relationships of the canonical analysis indicate how different instrumental orientations toward soap opera content lead to different viewing effects. The first canonical root reflected a highly involving orientation. Parasocial interaction, postviewing cognition, and postviewing discussion were related to viewing the serials to seek a variety of gratifications, including exciting entertainment, social utility, escapist relaxation, information, and voyeurism. This viewing pattern included affinity with a favorite soap opera, perceptions of realistic content, intentional planning to watch the program, and attention to the content when viewing. The second canonical root suggests an interesting contrast between interpersonal and parasocial interaction. Postviewing discussion about program content was linked to seeking social utility gratifications from serial viewing. The content need not be seen as realistic, nor must close attention be paid to the program when viewed. This is a more social and other-directed orientation toward soap opera use. The pattern is similar

19 264 HUMAN COMMUNICATION RESEARCH / Winter 1987 to one observed by Lemish (1985). In that study, some involved soap opera viewers were aware of the social context. These challengers derided the programs and characters in their conversations with others. On the other hand, parasocial interaction was clearly a sense of personal involvement with soap opera characters and stories. Parasocial interaction was linked to motives to seek voyeurism, information, and escapist relaxation gratifications from realistically perceived and attended to soap opera content. This is a more personal and innerdirected orientation toward soap opera use. These two contrasting orientations suggest that parasocial interaction may be a functional alternative to interpersonal interaction. The inner-directed orientation indicates, for example, that parasocial interaction with soap opera characters might be similar to readers involvement with romance novel heroines. Radway (1984) found that her informants read romance novels to escape from daily routines and to find private time, thought sexual content was appealing, felt they learned from the stories, saw the settings and characters as realistic, and became highly involved with the heroines. The readers did not evidence social or psychological deficiencies, but did derive personal satisfaction from the private world created by their reading. Using soap operas as a functional alternative by the inner-directed viewers of the present sample may be a relaxing way to escape more demanding interpersonal interaction at the time. A ritualistic use of soap operas was not identified in this study. Because of the focus on college students favorite soap operas, this should not be too surprising. Clearly, the appeal of a particular program that makes it an avid audience member s favorite is associated with more instrumental or goal-directed involvement. Minimal time availability for these respondents also may explain the more purposive orientation. Because media behavior also can be ritualistic, though, the results of this study should be interpreted with some caution. Activities and consequences associated with ritualistic media use may be of a different nature than the dimensions of involvement examined here. Perhaps, displacement effects, knowledge gaps, and dependency may result from using a medium less actively and more ritualistically (Windahl, 1981). Future research should consider such consequences. The findings also show how involvement might be seen in a more traditional sense as a precommunication condition. Certain elements that viewers brought with them to their viewing, such as attitudes about

20 Rubin, Perse / SOAP OPERA INVOLVEMENT 265 soap opera content and viewing motives, were significant correlates of involvement. But, soap opera involvement also was dependent on how intentionally respondents approached the viewing situation and how much attention they paid to the programs. It also is likely that participant involvement affects subsequent viewing motives, attitudes, and activity. Future research should consider how, for example, parasocial interaction with media personalities influences the use and effects of media content. This study, then, provides additional support for the variable and multidimensional nature of audience activity. As Windahl (1981) argued, although not superrational or universally selective, audience members are guided by expectations and perceptions of media content, which is another way of saying that the communication process to a large extent contains subjective choices and interpretations (p. 179). This investigation emphasizes the importance of including attitudinal, activity, and involvement measures in uses and effects inquiries. These factors increase our understanding of audience orientations toward the media and their content, as well as media effects. Future research should look beyond the affective, cognitive, and behavioral outcomes of this study to include dimensions of involvement as variables that intervene between audience uses and more traditional media effects. NOTES 1. The purposive sample was appropriate because the study was not designed for generalization to the population, but to investigate variable relationships (Wimmer & Dominick, 1987, p. 71). As Tucker, Weaver, and Berryman-Fink(l981, p. 162) reasoned, the ability to generalize widely from a single study is less than critical because behavioral science theory evolves from principles demonstrated on a large number of well designed and executed studies. College students are soap opera fans, purposive student samples have been widely used in this type of research, and several of the instruments have been used reliably with such samples. 2. The modified statements in the 10-item PSI (with means and standard deviations) were (a) my favorite soap opera character makes me feel comfortable, as if I am with a friend (2.18,1.07); (b) I see my favorite soap opera character as a natural, down-to-earth person (2.58,1.23); (c) I look forward to watching my favorite soap opera character on tomorrow s episode (2.70, 1.10); (d) if my favorite soap opera character appeared on another TV program, I would watch that program (2.72,l.ll); (e) my favorite soapopera character seems to understand the kinds of things I want to know (2.00,O.M); (f) if1 saw a story about my favorite soap opera character in a newspaper or magazine, I would read it (3.21,l.M); (9) I miss seeing my favorite soap opera character when he or she is ill or on vacation (2.37,1.16); (h) I would like to meet my favorite soap opera character in person

21 266 HUMAN COMMUNICATION RESEARCH / Winter 1987 (2.95,1.29); (i) I feel sorry for my favorite soap opera character when he or she makes a mistake (2.70,1.08); (j) I find my favorite soap opera character to be attractive (3.35,1.32). REFERENCES Amheim, R. (1944). The world of the daytime serial. In P. F. Lazarsfeld & F. N. Stanton (Eds.), Radio research (pp ). New York: Duel, Sloan & Pearce. Bailyn, L. (1959). Mass media and children: A study of exposure habits and cognitive effects. Psychological Monographs, 73(1). Barbatsis, G. (1983). Soap opera as etiquette book Advice for interpersonal relationships. Journal of American Culture, 6(3), Barwise, T. P., Ehrenberg,A.S.C., &Goodhardt, G. J. (1982). Glued to the box?patterns of TV repeat-viewing. Journal of Communicafion, 32(4), Bechtel, R. B., Achelpohl, C., & Akers, R. (1972). Correlates between observed behavior and questionnaire responses on television viewing. In E. A. Rubinstein, G. A. Comstock, & J. P. Murray (Eds.), Television and social behavior, uol. 4: Television in day-to-day life, patterns of use (DHEW Pub. HSM , pp ). Washington, DC: Government Printing Office. Blumler, J. G. (1979). The role of theory in uses and gratifications studies. Communication Research, 6,9-36. Bryant, J., & Comisky, P. W. (1978). The effect of positioning a message within differentially cognitively involving portions of a television segment on recall of the message. Human Communication Research, 5, Cantor, M. G., & Pingree, S. (1983). The soap opera. Newbuy Park, CA Sage. Capella, J. N. (1983). Conversational involvement: Approaching and avoiding others. In J. M. Wiemann & R. P. Harrison (Eds.), Nonwrbal interaction (pp ). Newbury Park, CA Sage. Carveth, R., & Alexander, A. (1985). Soap opera viewing motivations and the cultivation process. Journal of Broadcasting & Electronic Media, 29, Cassata, M. B. (1983). Plus Fa change, plus c est la m6me chose: An analysis of soap opera from radio to television. Journal of American Culture, 6(3), Fine, M. G. (1981). Soap opera conversations: The talk that binds. Journal of Communication, 31(3), Gantz, W. (1978). How uses and gratifications affect recall of television news. Journalism Quarterly, 55, ,681. Greenwald, A. G. (1968). Cognitive learning, cognitive response to persuasion, and attitude change. In A. G. Greenwald, T. C. Brock, & T. M. Ostrom (Eds.), Psycholqg cal foundations of attitudes (pp ). New York Academic Press. Greenwald, A. G., & Leavitt, C. (1984). Audience involvement in advertising: Four levels. Journal of Consumer Research, 11, Hill, D. B. (1985). Viewer characteristics and agenda setting by television news. Public Opinion Quarterly, 49, Horton, D., & Wohl, R. R. (1956). Mass communication and para-social interaction: Observations on intimacy at a distance. Psychiatry, 19, Katzman, N. (1972). Television soap operas: What s been going on anyway? Public Opinion Quarterly, 36, Kellermann, K. (1985). Memory processes in media effects. Communication Research, 12,

22 Rubin, Perse / SOAP OPERA INVOLVEMENT 267 Krugman, H. E. (1966). The measurement of advertising involvement. Public Opinion Quarterly, 30, 5834%. Lambert, Z. V., & Durand, R. M. (1975). Some precautions in using canonical analysis. Journal of Marketing Research, 12, Lemish, D. (1985). Soap opera viewing in college: A naturalistic inquiry. Journal of Broadcasting & Electronic Media, Levine, M. S. (1977). Canonical analysis and factor comparison. Newbury Park, CA Sage. Levy, M. R. (1978). The audience experience with television news. Journalism Monograph, 55. Levy, M. R. (1983). Conceptualizing and measuring aspects of audiepce activity. Journalism Quarterly, 60, Levy, M. R., & Wmdahl, S. (1984). Audience activity and gratifications: A conceptual clarification and exploration. Communication Research, 11, Levy, M. R., & Windahl, S. (1985). The concept of audience activity. In K. E. Rosengren, L. A. Wenner, & P. Palmgreen (Eds.), Media gratijications research: Current perspectives (pp ). Newbury Park, CA Sage. McDonald, F. (1974, October). CPMs and rating points aren t the whole story. Marketing and Media Decisions, pp. 6061,118,124. McGuire, W. J. (1974). Psychological motives and communication gratification. In J. G. Blumler & E. Katz (Eds.), The uses of mass communications: Current perspectiws on gratifications research (pp ). Newbury Park, CA: Sage. McLeod, J. M., & Becker, L. B. (1974). Testing the validity of gratification measures through political effects analysis. In J. G. Bumler & E. Katz (Eds.), The uses of mass communications: Current perspectives on gratifications research (pp ). Newbury Park, CA Sage. McLeod, J. M., & Becker, L. B. (1981). The uses and gratifications approach. In D. D. Nimmo & K. R. Sanders (Eds.), Handbook of political communication (pp ). Newbury Park, CA: Sage. Miyazaki, T. (1981). Housewives and daytime serials in Japan: A uses and gratifications perspective. Communication Research, 8, Palmgreen, P. (1984). Uses and gratifications: A theoretical perspective. In R. N. Bostrom (Ed.), Communication yearbook 8 (pp ). Newbury Park, CA Sage. Perse, E. M. (1986). Soap opera viewing patterns of college students and cultivation. Journal of Broadcasting & Electronic Media, 30, Petty, R. E., & Cacioppo, J. T. (1979). Issue involvement can increase or decrease persuasion by enhancing message-relevant cognitive responses. Journal of Personality and Social Psychology, 37, Petty, R. E., & Cacioppo, J. T. (1984). The effects of involvement on responses to argument quantity and quality: Central and peripheral routes to persuasion. Journal of Personality and Social Psychology, 46, Potter, W. J. (1986). Perceived reality and the cultivation hypothesis. Journal of Broadcasting & Electronic Media, 30, Radway, J. A. (1984). Reading the romance: Women, patriarchy, andpopular literature. Chapel HiU: University of North Carolina Press. Rose, B. (1979). Thickening the plot. Journal of Communication, 29(4), Rosengren, K. E., & Windahl, S. (1972). Mass media consumption as a functional alternative. In D. McQuail (Ed.), Sociology of mass communications (pp ). Middlesex, England Penguin.

23 268 HUMAN COMMUNICATION RESEARCH / Winter 1987 Rouner, D. (1984). Active television viewing and the cultivation hypothesis. Journalism Quarterly, 61, Rubin, A. M. (1981). An examination of television viewing motivations. Communication Research, 8, Rubin, A. M. (1983). Television uses and gratifications: The interactions of viewing patterns and motivations. Journal of Broadcasting, 27, Rubin, A. M. (1984). Ritualized and instrumental television viewing. Journal of Communication, 34(3), Rubin, A. M. (1985). Uses of daytime television soa~ operas by college students. Journal of Broadcasting & Electronic Media, 29, Rubin, A. M., & Perse, E. M. (1987). Audience activity and television news gratifications. Communication Research, 14, Rubin, A. M., Perse, E. M., &Powell, R. A. (1985). Loneliness, parasocial interaction,and local television news viewing. Human Communication Research, 12, Rubin, R. B., & McHugh, M. P. (1987). Development of parasocial interaction relationships. Journal of Broadcasting & Electronic Media, 31, Sherif, C. W., Sherif, M., & Nebergall, R. E. (1965). Attitude and attitude change: The social judgment-involvement approach. Philadelphia: Saunders. SPSS, Inc. (1986). SPSSX user s guide (2nd ed.). New York: McGraw-Hill. Stanford, S. W. (1984). Predicting favorite TV program gratifications from general orientations. Communication Research, 11, Tan, A. S. (1980). Mass media use, issue knowledge and political involvement. Public Opinion Quarterly, 44, Timberg, B. (1983). The rhetoric of the camera in television soap opera. Journal of American Culture, 6(3), Tucker, R. K., & Chase, L. J. (1980). Canonical correlation. In P. R. Monge & J. N. Cappella (Eds.), Multivariate techniques in human communication research (pp ). New York Academic Press. Tucker, R. K., Weaver, R. L., & Berryman-Fink, C. (1981). Research in speech communication. Englewd Cliffs, NJ: Prentict-Hall. Turow, J. (1974). Advising and ordering: Daytime, prime time. Journal of Communication, 24(2), Weaver, D. H. (1977). Political issues and voter need for orientation. In D. L. Shaw & M. E. McCombs (Eds.), The emergenceofamericon politicalissues: The agenda-setting function of the press (pp ). St. Paul: West. Wenner, L. A. (1983). Political news on television: A reconsideration of audience orientations. Western Journal of Speech Communication, 47, Whetmore, E. J., & Kielwasser, A. P. (1983). The soap opera audience speaks: A preliminary report. Journal of American Culture, 6(3), Wimmer, R. D., & Dominick, J. R. (1987). Mass media research: An introduction (2nd ed.). Belmont, CA: Wadsworth. Windahl, S. (1981). Uses and gratifications at the crossroads. In G. C. Wlhoit & H. de Bock (Eds.), Mass communication review yearbook (Vol. 2, pp ). Newbury Park, CA Sage.