éé’g WWWIWIHHJHHHHI\HIIHIHHIHWHHHWHI IIIIIIIIIIIIII IIIIIIIIIIIIIIIIIII 31293 00648 6033 III 3L LIBRARY Michigan State University This is to certify that the thesis entitled The effect of source likability and argument strength on cognitive processing: Do high and low self-monitors use these cues differently when processing persuasive messages? I presented by Richard J. Harnish has been accepted towards fulfillment of the requirements for Master of Arts degree in Psychology og/Mym Major professor Date Julv 29. 1987 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution MSU .— RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. 4/75” 777/ THE EFFECT OF SOURCE LIKABILITY AND ARGUMENT STRENGTH 0N COGNITIVE PROCESSING: DO HIGH AND LON SELF-MONITORS USE THESE CUES DIPPERENTLY HHEN PROCESSING PERSUASIVE MESSAGES? BY Richard J. Harnish A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Psychology 1987 ABSTRACT THE EFFECT OF SOURCE LIKABILITY ON COGNITIVE PROCESSING: DO HIGH AND LON SELF-HONITORS USE SOURCE LIKABILITY DIPFERENTLY HHSN PROCESSING PERSUASIVE MESSAGES? By Richard J. Harnish The research findings of DeBono (1986, 1987; Snyder & DeBono, 1985) suggest that high self-monitors--for whom attitudes serve a social- adjustive function-~would be more persuaded by a message that was 'I presented by a likable rather than a dislikable source and that low self-monitors--for whom attitudes serve a value-expressive function-- would be more persuaded by strong rather than weak message arguments. To test these hypotheses, undergraduate subjects differing in their self-oonitoring propensities read a strong or weak persuasive message written by a likable or dislikable source. Overall, results in general yielded partial support for the functional perspective but the pattern of effects was not as predicted. The results suggest that for high self-monitors, either source likability 9; strong message arguments were dif/ sufficient to produce greater attitude favorability, whereas for low i .._. “W M self-monitors, the predicted strength of argument effect occurred only guy.- ii ’: II I when the source was likable. ACKNOWLEDGEMENTS I am indebted to many people for my progress through graduate school. I wish to express my sincerest gratitude to my family (Alice, John, Stella, and Leslie) who always believed in my abilities and encouraged me to develop them to the fullest of my potential. Without your emotional and financial support my progress in graduate school and this thesis would not have been possible. Thank you with love. I also wish to express my gratitude to my committee for their guidance and support. Primary among these people is Chuck Stangor who believed in my abilities and offered encouragement. Thank you for being a good friend and advisor. I wish to also thank Bob Emmons who has been instrumental in this and other works. Many thanks also go to Ken DeBono for introducing me to functional theories and self-monitoring, and for helping me formulate my thoughts and ideas on this topic. I also wish to thank Larry Hesse for his attention to details and statistical expertise. Thank you for your time, effort, and guidance. Among others, I wish to thank Brad Warren and Leslie Cohen who were helpful in the data collection and coding. I also wish to thank Larry Marcy, Linda Sullivan, and Dave Sattler for being very good friends and offering their social support. Finally, I thank God for my very good fortune. ii TABLE OF CONTENTS List of Tables ___________________________________________ v Models of Information Processing ______________________________________ 1 Functional Theories of Attitudes _ ____ ___________ 3 Self-Monitoring and Message Arguments _______ _ __5 Source Characteristics _ _ _ _ __6 Self-Monitoring, Source Characteristics, and Message Arguments ___________________________ __ _ ___7 Hypotheses _____________________________________________________ 8 Method _______________________________________________________________ 11 Subjects and Design_ _ _______________________________ 11 Procedure: Independent Variables _____________________________ 12 Self-Monitoring ____________________________________________ 12 Sources ____________________________________________________ 12 Messages _ ________________________________________ 13 Strong Arguments ________________________________________ 13 Weak Arguments __________________________________________ 14 Dependent Measures ____________________________________________ 1S Attitude Measure ___________________________________________ 15 Cognitive Measures _________________________________________ 16 Manipulation Check _________________________________________ 16 iii Results ___________________________ _ ____________________ 17 Manipulation Check of Source Likability __ 17 Postmessage Attitude Scores __________ _-_ _ _ 17 Cognitive Response Analysis ___________________________________ 21 Total Thoughts Listed ______________________________________ 21 Thoughts in Support of Message _____________________________ 22 Counterarguments ___________________________________________ 25 Recall Analysis _______________________________________________ 28 Discussion _ _______________________________ 29 Cognitive Response Analysis ____________________________ 30 Recall Analysis ____________________________________ 31 Conclusions ___________________________________________________ 32 Appendix A: Pretest Materials _______________________________________ 35 Appendix B: Experimental Materials __________________________________ 46 List of References ______________ _ __ __ _60 Footnotes ____________________________________________________________ 63 iv LIST OF TABLES TABLE PAGE 1. Mean Postmessage Favorability/Unfavorability Scale Scores ______ 19 2. Mean Supportive Thoughts Listed ________________________________ 24 3. Mean Counterarguments Listed________‘ __________________________ 27 MODELS OF INFORMATION PROCESSING Recent investigations in the area of persuasion have focused on the modes of information processing that individuals adopt to ascertain the validity of persuasive messages. Two models have received the most attention in recent years. Attitude change theories have noted that the defining characteristic of the first model--which Chaiken (1980) has called "systematic," and Petty and Cacioppo (1981) have termed the "central route to persuasion"--is its emphasis on processing the content of the information contained in the message. In contrast, the second mode--labeled "heuristic“ processing by Chaiken (1980) or the ”peripheral route" by Petty and Cacioppo (1981)--is characterized by the use of tangential, less complex informational cues (e.g., communicator attractiveness) that, for whatever reasons, serve as "marker variables" for message validity. In particular, Chaiken (1980) has distinguished between these two processes. There are times when individuals actively process the content of a message; that is, when individuals think critically about the message. In this case, persons elaborate and extend the message arguments by relating them to personal experiences--in other words, they are responsive to the Quality of the message arguments. To the extent that such processing occurs, persuasion is thought to be a function of 1 2 the valence of the thoughts generated in response to the message. Favorable thoughts (i.e., thoughts in support of the position advocated) should enhance persuasion, whereas unfavorable thoughts (i.e., counterarguments) should inhibit persuasion. As noted, Chaiken refers to this mode as a "systematic" processing approach. In contrast, there are times when individuals are less thorough in evaluating the validity of a message. Instead of being systematic in the analysis of the message, individuals use simple decision rules related to non-message cues in the persuasion context. These decision rules, or heuristics, are beliefs about non-message factors (e.g., communicator trustworthiness) that are accepted as indices of information quality. Chaiken's (1980, 1982, 1986) heuristic model proposes that people often use simple decision rules when judging the validity of a persuasive message. For example, some of these simple decision rules are: "length implies strength," "experts can be trusted," and "consensus implies correctness." Without fully absorbing and processing the information presented, people might agree more with messages that contain many rather than few arguments, with expert rather than nonexpert communicators, or with messages with which many rather than few people agree (Chaiken, 1986). Studies have shown that source credibility, source likability, physical attractiveness, message length, number of arguments, audience reaction, and consensus information have their greatest impact on persuasion when people are not systematically processing message information (Axsom, Yates, & Chaiken, 1987; Cacioppo & Petty, 1984; Chaiken, 1980, 1986; Haugtvedt et al., 1986; Petty 8 Cacioppo, 1984; 3 Petty et al., 1981, 1983; Ratneshwar S Chaiken, 1986; flood, Kallgren, & Preisler, 1985; Yalch & Elmore-Yalch, 1984). In addition, recent research also has shown that a large number of variables either motivate or enable the individual to engage in systematic processing of a persuasive message, including the personal relevance of a message (e.g., Howard-Pitney et a1, 1986; Petty S Cacioppo, 1984), the match between a message's content and the recipient's functional predispositions (Cacioppo et al., 1982; DeBono, 1987), and amount of exposure to the message (Cacioppo 8 Petty, 1985). Indeed, research on motivational variables and their effect on cognitive processing has suggested that motivational factors have important and meaningful effects on a wide range of behaviors (Borgida, 8 Howard-Pituey, 1983; Erber, & Fiske, 1984; narkness, DeBono, S Borgida, 1985). Functional theories of attitudes As part of the recent increase in attention such motivational variables have been receiving, there has been a reawakening of interest in functional theories of attitudes (Katz, 1960; Smith, Bruner, & White, 1956). These theories assume that there are certain individualistic needs that are being met by one's attitudes, and that these attitudes allow the individual to implement certain plans to attain certain goals. Four functions, in particular, have been proposed: ego-defensive, attitudes formed to protect oneself from undesirable truths; knowledge (object appraisal), attitudes that are formed to give meaning to objects; value-ex ressive, attitudes that permit the individual to express his or her own beliefs or dispositions; and social-ad'ustive, attitudes that are formed on the basis of how well they permit the individual to fit into certain situations and permit him or her to 4 behave in a socially appropriate manner in regard to various reference groups (Katz, 1960; Smith et a1. 1956). Functional theories, by their definition, envision attitude change as a peripheral route process. That is, no extension or elaboration of ‘the quality of the arguments supporting an attitude is necessary for attitude change to take place. According to the functional approach, to bring about attitude change one only needs to demonstrate to an individual that his or her existing attitude is not optimally serving its function and that a different attitude would better serve the individual's needs. There does, however seem to be a fundamental difference between the functional approach to attitude change and the peripheral or heuristic route as defined within the information processing perspective. Previous research investigating the peripheral route has focused exclusively on persuasion that is mediated by non-message factors, for example, source expertise, source attractiveness, and source likability (Chaiken, 1980). Functional theories in contrast, by their very nature, involve the plans, goals, and needs of the individual. Previous research (e.g., Borgida, & Howard-Pitney, 1983; Erber, & Fiske, 1984; Harkness, DeBono, & Borgida, 1985) has indicated that information pertaining to one's plans, goals, and needs tends to motivate individuals to focus their attention on all relevant information in the immediate environment. That is, most peripheral cues by their very nature direct attention gggy from attitude-relevant arguments (but see Hood 8 Eagly, 1981), whereas, functional cues by their very nature may direct attention towards attitude-relevant arguments. 5 In a situation where both functional cues and attitude-relevant information are present, one might expect systematic processing of information to occur. That is, although attitude change could be brought about by a functional cue, individuals might also elaborate and extend the message-relevant arguments that have been presented. For example, if a person possesses an attitude on an issue that is serving a social-adjustive function, any information pertaining to the inappropriateness of the pre-existing attitude and the appropriateness of the new attitude for presenting oneself in a socially appropriate manner, in addition to facilitating attitude change, should also capture the attention of the individual. Further, in a functionally relevant context, the individual should systematically process any other incoming information concerning the new attitude (DeBono, 1987). Self-monitoring and message arguments Recently, DeBono (1986, 1987) and Snyder 8 DeBono (1985) have examined the functional approaches used by high and low self-monitoring individuals (Snyder, 1974) to ascertain the validity of a persuasive message. High self-monitors are individuals who regulate their expressive self-presentation for the sake of public appearance. These persons are highly responsive to social and interpersonal cues concerning situational appropriate behaviors. Moveover, high self- monitors are concerned with impression management type issues and therefore strive to be the "right person in the right place, at the right time" (Snyder S Gangestad, 1986). In contrast, individuals low in the personality construct of self- monitoring lack the ability or motivation to regulate their expressive self. Instead, their behaviors are thought to functionally reflect 6 their own enduring or momentary inner states--their own attitudes, traits, and feelings. 0f prime concern to these individuals is that their behaviors and internal states remain consistent across social situations (Snyder 8 Gangestad, 1986). In his investigations, DeBono (DeBono, 1986, 1987; Snyder & DeBono, 1985) has shown that attitudes serve primarily a social-adjustive function for high self-monitors. That is, high self-monitors experienced more attitude change after listening to a social-adjustive message. In contrast, DeBono found that attitudes serve primarily a value-expressive function for low self-monitors. That is, low self- monitors showed more attitude change after exposure to a value- exgressive message. Source characteristics DeBono's research was concerned with the content of persuasive messages and the functions that these attitudes could serve individuals differing in their self-monitoring propensities. As such, his findings are consistent with the speculation that high self-monitors might be especially responsive to the attractiveness that a source possesses, whereas low self-monitors might be especially responsive to the expertise a source possesses. That is, high self-monitors could be especially responsive to a source that permits their attitudes to serve a social-adjustive function and, thus, such persons would be likely to perceive positions advocated by an attractive source as helpful in achieving their goal to fit into important social and interpersonal situations. In contrast, low self-monitors could be especially responsive to an expert source because the source might permit their attitudes to serve a 7 value-expressive function. That is, the attitudes presented by an expert source might help low self-monitors express their true self--in other words, their underlying values, beliefs, and dispositions if the low self-monitor holds the same attitude. By agreeing with the expert source, low self-monitors could reaffirm their own values and remain true to self. Self-monitoring, source characteristics, and message arguments The purpose of the present research was to examine if source characteristics (i.e., likability), and argument quality (i.e., strength) might serve different functions for high and low self- monitors. Because the high self-monitor regulates his or her expressive self for the sake of public appearance, any source that is considered important--that is, that could be useful to the high self-monitoring individual as a means for achieving the goal of presenting him or her self as behaving in a socially appropriate manner and thus being socially desirable--shou1d motivate high self-monitors to expend cognitive energy to process messages presented by an attractive source thoroughly. In contrast, because low self-monitors strive to have their behaviors and internal states remain consistent across social situations, message quality could be useful for the low self-monitoring individual in that strong compelling arguments might permit low self- monitors to appear consistent. Thus, low self-monitors would be more attentive to the quality of a persuasive message in general and be motivated to expend cognitive energy to process strong messages arguments. 8 In this study, three dependent measures were used to test the notion that source likability and message quality would serve different functions (e.g., social-adjustive or value-expressive) for high and low self-monitors. An attitude measure was used to measure postmessage belief change and two other procedures were used to assess cognitive effort. The first procedure was a cognitive response measure (Brock, 1967; Greenwald, 1968) that examined subjects' thoughts. It is believed that the mode of processing (i.e., systematic or heuristic) is reflected in the nature of the thoughts that occur to individuals while they listen to the message. A high proportion of message-relevant thoughts (e.g., arguments in support of the position advocated, counterarguments, and elaborations and extensions of the message arguments) is believed to be evidence of a systematic approach. The second cognitive measure used was an assessment of recall. It is believed that the more an individual thinks about and elaborates a message (i.e., processes it systematically), the more likely it is that the message will be stored in long term memory, and thus, be recalled correctly later (Craik 8 Lockhart, 1972). Hypotheses Hypothesis 1a: Given the results of previous research on persuasion and source likability (Sampson 8 Insko, 1964), a main effect for source likability was expected. Thus, persuasion, as measured by the postmessage attitude scale, should be greater when the source was likable rather than dislikable. Hypothesis 1b: In addition, the more likable the source the more attention subjects might give to that person and in turn, they might be more sensitive to the quality of the persuasive message. That is, the more likable an individual is the more 9 captivating or interesting that person might be. This increased awareness might motivate a person to systematically process a speaker's persuasive message. It was expected, therefore, that subjects would generate more supportive arguments when presented with a likable source, and that subjects would generate more counterarguments when presented with a dislikable source (as measured by the cognitive response measure). Hypothesis 1c: The expected greater motivation to systematically process information presented by a likable source, along with the elaborations (i.e., the generation of supportive arguments) of the persuasive message presented by a likable source, should facilitate the recall of message arguments (Craik 8 Lockhart, 1972). Thus, it was predicted that subjects would more accurately recall arguments that had been presented by a likable rather than a dislikable source. Hypothesis 2a: A main effect for the quality of the argument was also predicted. It was expected that subjects would be more persuaded by strong rather than weak arguments (as measured by the postmessage attitude scale). Hypothesis 2b: It was also expected that subjects would generate more supportive thoughts when presented with the strong arguments, but generate more counterarguments when presented with the weak arguments (as measured by the cognitive response measure). Hypothesis 2c: Further, the strong arguments should be recalled more accurately than the weak arguments, because the individual should have thought about and supported the strong arguments to a greater extent. Hypothesis 3a: Also predicted was a two-way interaction between self-monitoring and source likability. It was expected that high self- monitors would be more persuaded by a likable source (as measured by the postmessage attitude scale) because the likable source would serve a 10 social-adjustive function for the high self-monitor. Because source likability should not be as useful a function for low self-monitors, it was expected that low self-monitors would not be as affected by the varied degree of source likability. Hypothesis 3b: Because of the increased attention high self-monitors should give to a likable source, high self-monitors, rather than low self-monitors should generate more supportive arguments (elaborations and extensions) for the likable source's position, in comparison to a dislikable source's position, as measured by the cognitive response measure. In addition, it was also expected that high self-monitors, in comparison to low self-monitors, would generate more counterarguments for a dislikable rather than a likable source's position. Hypothesis 3c: Because of the elaborations and extensions made by the high self-monitoring subjects to the message arguments which were presented by a likable source, high self-monitors, in comparison to low self-monitors, should recall more accurately the message arguments which were presented by the likable source (as measured by the recall measure). Hypothesis 4a: It was expected that because source characteristics should not serve as useful a function for low self- monitors, these individuals would be more attentive to the quality of a persuasive message in general because of their concern that their behaviors and internal states remain consistent across social situations. Thus, message quality should serve a value-expressive function for low self-monitors allowing them to possess only those attitudes that are congruent with their behaviors. It was predicted that low self-monitors, in comparison to high self-monitors, would be more persuaded by strong rather than weak arguments (as measured by the 11 postmessage attitude scale). Hypothesis 4b: Further, low self- monitors, rather than high self-monitors, should be motivated to systematically process the strong message arguments and to generate more supportive thoughts for the strong arguments (as measured by the cognitive response measure). In addition, it was also predicted that low self-monitors, in comparison to high self-monitors, would generate more counterarguments for the weak rather than strong arguments. Hypothesis 4c: Low self-monitors then, should recall more message arguments (as measured by the recall measure) when the arguments were strong, since low self-monitors would have elaborated and extended the strong message arguments. Because of the elaboration, there would be a greater chance of the message being stored in long term memory (Craik 8 Lockhart, 1972). Method Subjects and Design One hundred and three undergraduates (24 males, 79 females) participated in this study to earn extra credit towards their grade in their introductory psychology course.1 Subjects participated in small, mixed-sex groups of approximately 15 people each. Based on their responses to pretest materials (i.e., the Self-Monitoring Scale, Snyder 8 Gangestad, 1986, and a celebrity likability questionnaire), high or low self-monitoring subjects were randomly assigned to the experimental conditions of a 2 (source of message; likable or dislikable) x 2 (argument strength; strong or weak) factorial design. 12 Procedure; Independent Variables Self-monitoring As part of a larger questionnaire study administered earlier in the term, individuals completed the abbreviated version of the Self- Monitoring Scale (Snyder 8 Gangestad, 1986). This new version consists of the 18 items of the original 25-item measure, (Snyder, 1974), that most validly assesses the general self-monitoring factor. 0n the basis of a median split (median = 10.25) of their responses to the Self- Monitoring Scale, half the participant were classified as high self- monitors (scores 1 10) and half as low self-monitors (scores 3 9). Sources Individuals also completed a questionnaire which examined celebrity likability. This questionnaire consisted of television, cinema, and recording artists. Subjects indicated on a 7-point Likert scale (1 = not at all, 7 = very much) the extent to which they liked the following celebrities: Bill Cosby, Bruce Willis, Johnny Carson, David Letterman, Barry Manilow, David Lee Roth, Prince, Stevie Wonder, Howard Cosell, Bruce Jenner, Glenn Close, Carol Burnett, Valerie Harper, Jane Fonda, Meryl Streep, Barbara Streisand, Madonna, Brooke Shields, Joan Rivers, Diana Ross, and Joan Collins. Because a large majority of subjects indicated that they liked Bill Cosby and Bruce Willis, and disliked David Lee Roth and Barry Manilow, those subjects who indicated that they liked (minimum rating = 5) Bill Cosby or Bruce Willis, or indicated that they disliked David Lee Roth or Barry Manilow (maximum rating = 4) on the prescreening celebrity attitude questionnaire were selected for the study. 13 5.932992 Subjects were told, "A Detroit-based consulting firm, Webster, Webster 8 Maxell, has asked us to help them with some market research. They are interested in knowing if there is any student support for a charity event that would raise money for the hungry and homeless. The event would take place later this fall. They have already contacted a few celebrities, but some of the celebrities were not sure whether or not to participate in the fund raiser. In recruiting the celebrities, the marketing firm said, 'Lately there has been much concern in the entertainment community to help those Americans and other Nationals that are hungry and homeless. Only months ago, Hands Across America took place to raise funds for the hungry and homeless in America. Later this fall, Home Aid will take place. Home Aid is a group of entertainers, such as movie and televisions stars, theatrical performers, and recording artists. Could we count on your support by giving a short performance?'” "One of the celebrities contacted was This was his reaction to the firm's request. Please take the next two minutes to read the statement." Following the introduction, subjects then read one of two sets of arguments why they should not support Home Aid consisting of either strong, compelling arguments, or weak, specious arguments. Strong arguments. The strong arguments contained in the message were as follows: "Although I can see why people might view Home Aid as a good idea, I believe that Home Aid is a rather poor idea. Let me tell you of some of the reasons why I feel this way. First, the concerts are too commercialized and cost too much to put on. The publicity and hype l4 surrounding the concert is extensive. The promoters of the concert must pay the advertising agencies for their services with the money raised for the needy. In addition, stage crews, managers, and directors must be paid. Further, there is a rental charge on the stadium where the event takes place. Clearly, it appears that the costs of the production is too great. Secondly, after all the expenses have been paid, there's no real way of knowing if all of the remaining money actually gets to the needy. The money must be channeled into governmental and private agencies that have been set up to help the needy. However, because of the bureaucracy characterizing these agencies, embezzlement of these funds is accomplished rather easily. Lastly, and perhaps most importantly, people will think that the plight of the hungry and homeless is solved. At best, the funds that do get to the needy provide a temporary solution to a long term problem. We won't be able to cure the problem with concerts. Much more needs to be done to help the hungry and homeless. Because of these reasons, I urge you not to support Home Aid." Weak arguments. The weak arguments contained in the message were as follows: "Although I can see why people might view Home Aid as a good idea, I believe that Home Aid is a rather poor idea. Let me tell you of some of the reasons why I feel this way. First by supporting Home Aid, you'll be giving the promoters of the event the wrong impression that peeple are not tired of these carbon copy concerts. We've already experienced Live Aid, Band Aid, and Farm Aid--do we have to experience Home Aid too? By not supporting Home Aid, you'll be showing the 15 promoters and organizers that we what a new, fresh approach. These concerts all seem alike., Secondly, there's too much repetitive music in the concerts and not enough of a variety of performers. The promoters and organizers only invite top 40 bands to play. After you've listen to a few songs, you've heard them all. We need a wide range of artists and types of music to show that everyone cares about the needy, not just a few. Lastly, the concert only allots its performers a given amount of time to perform. After watching and waiting for hours to see and hear your favorite artist, five to ten minutes doesn't seem long enough. I'd rather go to a concert where my favorite performer is playing and make a donation there. Because of these reasons, I urge you not to support Home Aid." To determine the strength of the two sets of message arguments, a t-test was conducted on a pretest measure of perceived argument quality given to 42 subjects from the same population, but who did not participate in the main experiment. As expected, the two set of message arguments differed significantly, p(40) = 2.92, p < .006 (strong argument, 5 = 4.24, and the weak arguments, ! = 2.90). Depgndent Measures Attitude measure. The postmessage measure of persuasion was taken immediately following the message presentation. Subjects indicated their attitudes towards Home Aid on a 7-point Likert scale, where 1 = not at all, and 7 = very much so. The scale consisted of the following adjectives which were chosen on the basis of their degree of applicability to Home Aid: good, valuable, cruel,2 justified, needed, and worthy. The six items were collapsed into a single attitude measure 16 by taking a mean across the six items. Cronbach's alpha calculated for the attitude scale was .82. Cognitive measures. As noted, two dependent measures were used to examine the extent to which individuals systematically processed the message arguments. The first was a cognitive response analysis (Brock, 1967; Greenwald, 1968) which was administered after the postmessage attitude measure. Subjects were instructed to write down all the thoughts, ideas, and associations that they had about the statement as they read through the message. Once this task was completed, subjects were then told to go back through the thoughts that they had just listed and put a plus next to all thoughts that they felt were in support of the writer's position, a minus next to all thoughts that did not support the writer's position (i.e., counterarguments), and a zero next to all thoughts that were neutral with respect to the writer's position. When everyone finished, the experimenter asked subjects if, because there was extra time in the session, they would help us collect some data for another study. Everyone agreed, and subjects were asked questions about the probability of certain events occurring. This task took approximately 7 minutes. After the completion of the filler tasks, the second dependent measure--to assess recall of the message arguments--was administered. Subjects were told to recall as many of the arguments favoring non- support of Home Aid as they possibly could. Manipulation check. After the cognitive response analysis was completed, subjects responded to a short questionnaire that examined their feelings towards the source of the message. Selected item from this measure included: Did you know who the source of the message was? 17 (yes, no), and based on 7-point Likert scales, to what extent do you liked the source? (1 = got at a; , 7 = very much , and to what extent do you agree with the message? (1 = not at all, 7 = very much). Results Manipulation check of source likability An analysis of variance was conducted on the actors nested within source to examine any differences in the degree of likability among the actors. Results indicated that there was no difference between the two likable sources (i.e., Bill Cosby, fl = 6.68, and Bruce Willis, M = 6.52) in their degree of likability and that there was no difference in the degree of likability between the two dislikable sources (i.e., David Lee Roth, 5 = 3.30, and Barry Manilow, ! = 3.28), £32, 98) = .10, p > .90. Because the analysis revealed that there was no significant difference in the degree of likability between the two likable sources, or between the two dislikable sources, the specific source was disregarded in all subsequent analyses. The preliminary analysis also indicated that there was a significant difference in the degree of likability between the two types of sources (i.e., likable and dislikable), £31,98) = 834.14, p < .001, a result that is consistent with the intended manipulation. Postmessage attitude scores. Attitude scores, as measured by the postmessage attitude scale, were submitted to a 2 (self-monitoring) x 2 (argument strength) x 2 (source likability) ANOVA. The mean postmessage attitude scores, which indicate favorable-unfavorable attitudes, are presented in Table 1. (Items were keyed so that lower mean scores are indicative of more favorable attitudes toward Home Aid). The analysis 18 did not yield support for Hypothesis 1a, which predicted a main effect for source likability, 231,93) = 1.36, p < .25. This finding suggests that subjects were not differentially persuaded by a likable source (5 = 5.01) or a dislikable source (5 = 5.27), however the means were in the predicted direction. The AWOVA did reveal a marginal main effect for argument strength, {(1,93) = 2.70, p < .10. As predicted by H othesis 2a, more persuasion was experienced by subjects when they were exposed to the strong arguments (! = 4.99) than when they saw the weak arguments (! = 5.26). 19 Table 1 Mean Postmessage Favorability/Unfavorability Scale Scores ---_-v----‘----‘—-’----- w _— - --—----——-----“----—-—-— -------— --------------------—------—----—--—----------—-----—------—----- Likable Dislikable Likable Dislikable Argument strength Strong 5 06a 5.04s 4.59c 5 21d n = (13) (17) (11) (13) Weak 4.93s 5.81b 5.44d 5.00d O.--—‘----UI_.0--fl-----—--------__------------------------------------------ means with differing subscripts are significantly different at p < .05. 20 The analysis, however, did not yield support for H othesis 3a, which predicted a two-way interaction between self-monitoring and source likability, £(1,93) = .578, p > .45. This finding suggests that high and low self-monitors were not differentially persuaded by a likable source (5 8 5.00; p = 5.05, respectively) or a dislikable source (5 = 5.31; p = 5.10, respectively). In addition, H othesis 4a, which predicted a two-way interaction between self-monitoring and argument strength, also was not supported, £(1,93) = .00, p > .98. This finding suggests that there was no significant difference in the amount of persuasion experienced by high and low self-monitors when they heard strong arguments (fl = 5.05; M = 4.92, respectively) or weak arguments (fl = 5.31; p = 5.22, respectively). However, the analysis did reveal a significant three-way interaction between self-monitoring, argument strength, and source likability, £(1,93) = 5.89, p < .008, suggesting that the combined effect of argument strength, and source likability differentially affected the amount of persuasion in high and low self-monitoring individuals. Simple effects test (Keppel, 1982) indicated that it was low self-monitors who experienced more persuasion when the source was likable and presented strong arguments, contrary to H othesis 4a, which proposed that source likability should not serve a useful function for low self-monitors. Simple effects tests (Keppel, 1982) also indicated that high self-monitors who received a dislikable source that presented weak arguments, were the least persuaded--contrary to Hypothesis 3a, which proposed that argument strength would not serve a useful function for these individuals. 21 Overall, these results seem to indicate that for high self- monitors, either source likability pp strong message arguments were sufficient to produce greater attitude favorability but that the combined effects of pgpp source likability ppg strong message arguments did not generate more favorable attitudes than did the presence of one of these persuasive forces alone. That is, source likability and argument strength did not have an additive effect on the amount of persuasion reported by high self-monitors. In addition, these results seem to indicate that for low self-monitors, the predicted strength of argument effect occurs only when the source was likable. This result suggests that the combined effects of strong message arguments gpg source likability were sufficient to produce greater attitude favorability in low self-monitors. Cognitive Response Analysis Total thoughts listed. To examine the types of thoughts (i.e., supportive, counterargumentative, neutral) as determined by subjects' evaluations, the total thoughts listed were submitted to a 2 (self- monitoring) x 2 (argument strength) x 3 (type of thought; supportive, counterargumentative, neutral) AHOVA with repeated measures on the last factor. The analysis revealed a main effect for type of thoughts listed §(2,188) = 5.04, p < .007, suggesting that subjects listed more negative thoughts (5 = 3.37, SQ = 2.52) than positive thoughts (M = 2.62, §Q = 2.11) or neutral thoughts (5 = 2.33, SQ = 2.00). The ANOVA also yielded a significant two-way interaction between source likability and type of thought, (3 = 6.41, p < .002), as predicted by Hypothesis 1b. It was expected that subjects would generate more supportive arguments when presented with a likable source and generate more counterarguments when 22 presented with a dislikable source. That is, subjects generated more supportive arguments when the source was likable (fl = 3.24) than when the source was dislikable (p = 1.96). In addition, Hypothesis 2b predicted that subjects would generate more supportive thoughts when presented with strong rather than weak arguments and produce more counterarguments when presented with weak rather than strong arguments. The analysis, however, generated a nonsignificant two-way interaction between argument strength and type of thought listed, (3 = .07, p > .93), suggesting that subjects were not differentially affected by argument strength when generating supportive or counterarguments. The analysis did reveal a significant four-way interaction between self-monitoring, argument strength, source likability and thoughts listed, £32, 188) = 3.00, p =.05, suggesting that the combined effects of self-monitoring, argument strength, and source likability had an impact on the types of thoughts generated by the subjects. To further explore this interaction, the supportive thoughts, and counterarguments were then separately examined by a 2 (self-monitoring) x 2 (argument strength) x 2 (source likability) AWOVA. Thoughts in support of message. Hypothesis 3b, predicted that high self-monitoring individuals would be more likely to think about and elaborate the arguments presented by a likable (versus a dislikable) source. In addition, Hypothesis 4b, predicted that low self-monitors would be more likely to think about and elaborate the strong rather than weak arguments. To test these predictions, a 2 (self-monitoring) x 2 (argument strength) x 2 (source likability) ANOVA was conducted on the supportive thoughts listed by subjects. As seen in Table 2, the 23 analysis revealed a significant two-way interaction between self- monitoring and source, fi(1,94) = 4.3, p < .04. 24 Table 2 Mean Supportive Thoughts Listed *_—~~-‘—-----_--------“‘_-~-- Likable Dislikable Likable Dislikable Argument strength Strong 3.23 2.00 3.00 2.31 n = (13) (17) (11) (13) Week 4.08 1.00 2.64 2.54 n = (12) ( 9) (14) (13) Note. Lower means indicate less message consistent attitudes. 25 In support of Hypothesis 3b, post hoc analysis using the Tukey test (Keppel, 1982) further indicated that high self-monitors listed more supportive thoughts when the source was likable (M = 3.64) than when the source was dislikable (M = 1.65) p < .05. There was no significant difference in the amount of supportive thoughts listed by low self- monitors when the source was likable (M,= 2.80) or when the source was dislikable (M = 2.42) ns. The analysis also revealed a nonsignificant two-way interaction between self-monitoring and argument strength, a pattern which was predicted by Hypothesis 4b. That is, it was expected that low self-monitors would be more likely to generate supportive thoughts when presented with the strong rather than weak arguments, but this hypothesis was not supported, fi(1,94) = 0.0, p > .99. Counterarguments. As predicted by H othesis 3b, it was expected that high self-monitors would generate more counterarguments for a dislikable (versus a likable) source's position. In addition, Hypothesis 4b, predicted that low self-monitors would generate more counterarguments for the weak (versus strong) arguments. To test these predictions, a 2 (self-monitoring) x 2 (argument strength) x 2 (source likability) ANOVA was conducted on the counterarguments listed by subjects. The analysis revealed a marginally significant two-way interaction between self-monitoring and source, £(1,94) = 2.75, p < .10. In support of Hypothesis 3b, post hoc analysis using the Tukey test (Keppel, 1982) further indicated that high self-monitors listed more counterarguments when the source was dislikable (M = 4.65) than when the source was likable (M = 2.76) p < .05. Further, as expected there were no significant differences in the amount of counterarguments listed by low self-monitors when the source was dislikable (M = 2.88) versus when 26 the source was likable (M,= 3.15) p > .05. The mean number of counterarguments listed by high and low self-monitors are presented in Table 3. 27 Table 3 Mean Counterarguments Listed Source received Likable Dislikable Likable Dislikable Argument strength Strong 3.69 4 41 2 55 3 46 n = (13) (17) (11) (13) Weak 1.75 5.11 3.14 2.85 Note. Lower means indicate more message consistent attitudes. 28 The analysis also indicated that H othesis 4b, which predicted that low self-monitors, in comparison to high self-monitors, would generate more counterarguments for the weak rather than strong arguments, was not supported, {(1,94) = .40, p > .52. Recall analysis A main effect was predicted for source likability and argument strength on the recall measure. Hypothesis 1c predicted that subjects would more accurately recall arguments that had been presented by a likable rather than a dislikable source. In addition, Hypothesis 2c predicted that the strong arguments should be recalled more accurately than the weak arguments, because the individual should have thought about and supported the strong arguments to a greater extent. An interaction between self-monitoring and source likability was expected on the recall measure as well. Hypothesis 3c predicted that high self- monitoring subjects would recall more accurately the message arguments presented by a likable (versus a dislikable) source, since source likability should be serving a social-adjustive function for these individuals. In contrast, Hypothesis 4c predicted that low self- monitors would recall more accurately the strong (versus weak) arguments because argument strength should be serving a value-expressive function for these individuals. To examine these relationships, two judges coded the recalled message arguments to determine how accurately subjects recalled the persuasive message. Interrater reliability between the judges was .97. A 2 (self-monitoring) x 2 (argument strength) x 2 (source likability) ANOVA was then conducted on the recall measure. The analysis revealed no significant effects, all ps > .30. 29 Discussion The results of this study suggest that the combined effects of self-monitoring, source likability, and argument strength affected individuals postmessage reactions to Home Aid. Analysis of the postmessage attitude scale revealed that low self-monitors, in comparison to high self-monitors, were most persuaded by a likable source that presented strong arguments. Further, this analysis indicated that for high self-monitors, either source likability pp strong message arguments was sufficient to produce greater attitude favorability but that the combined effect of both source likability and strong message arguments did not generate more favorable attitudes. Analysis of the cognitive response analysis indicated that the supportive thoughts listed by high self-monitoring subjects was a function of the source's likability. That is, high self-monitors, in comparison to low self-monitors, listed more supportive thoughts when the source was likable than dislikable. Moreover, high self-monitors listed more counterarguments when the source was dislikable rather than likable. The results suggest that low self-monitors were not differentially affected by source likability (or by argument strength) in terms of the amount of supportive or counterarguments that these individuals listed. Overall, the results of this study generally support the hypotheses that peripheral cues and message quality have different effects as a function of self-monitoring. These findings have several implications 30 especially for researchers investigating attitudes and persuasion processes. It appears that source characteristics and argument strength have functionally different effects on high and low self-monitoring individuals in the amount of attitude change experienced and in the mode of cognitive processing performed on a persuasive message. Further, the attitude change brought about for high and low self-monitoring individuals appear to result from the differences in the type of function the attitude is serving. Just as high and low self-monitoring individuals' behavioral choices are reflected by either exterior or interior cues, so too are their choices in the amount of attention given to different aspects of a persuasive message. That is, because high self-monitors are sensitive and more responsive to cues in the environment, fewer cues are sufficient to produce persuasion and attitude change for these individuals. Low self-monitors, on the other hand, are less sensitive and responsive to cues in the environment and, thus, require more cues or a combination of cues to produce persuasion or attitude change. Cognitive response analysis High self-monitors listed more supportive thoughts in favor of a likable source's position, and they listed more counterarguments when the source was dislikable. In comparison, low self-monitors were not differentially affected by source likability in the amount of supportive thoughts or counterarguments that they listed. These findings seem to further reflect the findings of the postmessage favorability scale scores in that high self-monitors appeared to be most susceptible to persuasion by a single cue--source likability, whereas, low self- monitors were not influenced by source likability alone. This finding 31 only further emphasizes the ability high self-monitors have of shaping their behavior to external cues in the environment and the insensitivity or inability low self-monitors have molding their behavior to these cues. Because of the high premium high self-monitor place on appearing to be the right person in the right place in the right time, costs are incurred with this behavioral strategy in that the most readily available cue are used to determine how to process a persuasive message. Employing this strategy may lead high self-monitors to behavioral choices which they may regret later, after having acted on the persuasive message, and when the consequences of their actions are fully realized. Because of the importance low self-monitors place on having their behaviors remain consistent with their own feeling, beliefs, and opinions, low self-monitors seem to need additional cues for persuasion to occur. Employing this strategy incurs costs also, in terms of time and effort to process the additional cues, however, this strategy appears to be less costly to a low self-monitoring individual's sense of self. That is, this strategy permits low self-monitors to make behavioral choices that appear to be rather consistent with their previously held feelings, beliefs, and opinions and, thus, they may not later regret their behavioral choices when their actions are fully realized. Recall analysis From the number of thoughts generated in the cognitive response analysis (i.e., number of thoughts in support of the message, and number of counterarguments, and elaborations and extensions of the issue relevant thoughts), it appeared that at times systematic processing of the message may have occurred. Moreover, to the extent to which such 32 processing occurred, it was likely that the message was placed in long term memory (Craik 8 Lockhart, 1972), and the more accurate should have been the recall of the message arguments. Thus, it was expected that recall would be facilitated when attitudes were serving a function for individuals. However, the analysis revealed no significant results. One can only speculate why this failure occurred. Perhaps, because subjects were participating in the experiment to earn extra credit, they might have been motivated only enough to have the message stored in short term memory. After all, the message that the individuals received had little significance to them after they left the experimental session. Further, subjects were given no choice in determining whom to listen to, in contrast to real life where one is presented with the opportunity to selectively ”tune out" unwanted information or to ignore certain sources. Given these constraints and the artificiality of the study (i.e., subjects were told so and so wrote the message in response to a request to appear at a fund raiser) it is somewhat surprising that the experimental manipulations had such an impact as they did. Perhaps, in real life, with live sources and with topics which are more important to an individual, the probability of a persuasive message being stored in long term memory, and thus recalled at a later time, may be much greater. Conclusions While the present study is similar to DeBono's previous works in terms of its method and procedure, an important difference in terms of the present study's generalizability overshadows these similarities. Unquestionably, DeBono's research has shown that attitudes do in fact serve different functions for high and low self-monitors, however in 33 order to demonstrate this phenomena, DeBono had to construct sources and messages that would address the differing attitude functions of high and low self-monitors. Outside the laboratory, one may find that persuasive sources and persuasive messages are not so neatly categorized to solely address a social-adjustive or a value-expressive function for an individual. The present research has attempted to address this issue by examining source likability and message quality without attempting to construct highly specific functional cues to engender attitude change. How then, could one test functional attitudes without first presenting highly specific cues (e.g., introducing the speaker as a renown expert in his or her field of study) to individuals which would address their specific plans, goals, or needs? Source likability especially seemed to be an ideal construct in that one may form his or her evaluative judgment of a person immediately upon first encountering the person. Under such circumstances, the individual possesses little information about the person, and may not be able to determine if the person is an expert in a particular field or if the person possesses a high degree of status. However, the knowledge that the individual does possess about the person in question will be sufficient to determine if he or she will like that person. Given the results of this study--that there was not an additive effect of source likability and message quality for persuasion in high self-monitors and that persuasion only occurred for low self-monitors when the likable source presented strong arguments--likability may be naturally serving different functions for high and low self-monitors. The next question that now begs to be asked is, if source likability serves different functions for high and low self-monitors, 34 what makes a source likable for these individuals. The findings of this study coupled with those of DeBono suggest that for high self-monitors, likability may center around status-related concerns, such as, what can the likable source do for the high self-monitoring individual in terms of his or her plans, goals, and needs. In contrast, for low self- monitors, likability may revolve around the belief that the low self- monitoring individual and the likable source share a similar disposition and similar opinions. For instance, in the present study, both high and low self-monitors reported liking Bill Cosby, however what is not known is did these individuals like him for the same reasons? Did high self- monitors report liking Bill Cosby because he is a celebrity and for all the trappings that accompany fame? Did low self-monitors report liking Cosby because of his congenial personality? Future research should address these questions by further examining what makes a likable source likable for high and low self-monitoring individuals. By examining such naturally occurring functions as source likability, we will come to a further understanding of non-message factors and ultimately, what makes a persuasive message persuasive. APPENDICES APPENDIX A: Pretesting Measures 35 APPENDIX A: Pretesting Measures We would like your name and local phone number so that we could contact you to see if you're interested in participating in a few experiments that we'll be conducting this term. If you are interested in earning extra credit towards your grade in introductory psychology by participating in experiments, please complete the following: RACE: NHITE BLACK OTHER (PLEASE CIRCLE ONE) Your responses to the questionnaires and this personal information will be kept strictly confidential. 36 APPENDIX A: Pretesting Measures Attitudes Towards Celebrities Please indicate the extent to which you like the following celebrities by circling a number. For example, by circling the 4, it means that you feel neutral towards the celebrity. 1. Bill Cosby 1 2 dislike very much 2. Bruce Willis 1 2 dislike very much 3. Johnny Carson 1 2 dislike very much 4. David Letterman 1 2 dislike very much 5. Barry Manilow 1 2 dislike very much 7 like very much 7 like very much 7 like very much 7 like very much 7 like very much 6. Michael Jackson 1 2 dislike very much 7. David Lee Roth 1 2 dislike very much 8. Prince 1 2 dislike very much 9. Stevie Wonder 1 2 dislike very much 10. Howard Cosell 1 2 dislike very much 11. Bruce Jenner 1 2 dislike very much 12. Glenn Close 1 2 dislike very much 37 7 like very much 7 like very much 7 like very much 7 like very much 7 like very much 7 like very much 7 like. very much 38 13. Carol Burnett 1 2 3 4 dislike very much 14. Valerie Harper 1 2 3 4 dislike very much 15. Jane Fonda 1 2 3 4 dislike very much 16. Meryl Streep 1 2 3 4 dislike very much 17. Barbara Streisand 1 2 3 4 dislike very much 18. Madonna 1 2 3 4 dislike very much 19. Brooke Shields 1 2 3 4 dislike very much 20. Joan Rivers 1 2 3 4 dislike very much 7 like very much 7 like very much 7 like very much 7 like very much 7 like very much 7 like very much 7 like very much 7 like very much 21. Diana Ross 1 dislike very much 22. Joan Collins 1 dislike very much 2 39 7 like very much 7 like very much 40 APPENDIX A: Pretesting Measures Personal Reaction Inventory The statements on the following pages concern your personal reactions to a number of different situations. No two statements are exactly alike so consider each statement carefully before answering. If a statement is true or most true as applied to you, fill in the "T" next to the statement. If the statement is false or mostly false as applied to you, fill in the ”F" next to the statement. (I) (F) 1. I find it hard to imitate the behavior of other people. (T) (F) 2. At parties and social gatherings, I do not attempt to do or say things that others will like. (T) (F) 3. I can only argue for ideas which I already believe. (T) (F) 4. I can make impromptu speeches even on topics about which I have almost no information. (T) (F) 5. I guess I put on a show to impress or entertain others. (T) (F) 6. I would probably make a good actor. (T) (F) 7. In a group of people, I am rarely the center of attention. (T) (F) 8. In different situations and with different people, I often act like very different people. (T) (F) 9. I am not particularly good at making people like me. (I) (F) 10. I'm not always the person I appear to be. (T) (F) 11. I would not change my opinions (or the way I do things) in order to please someone or to win their favor. (T) (F) 12. I have considered being an entertainer. (T) (F) 13. I have never been good at games like charades or improvisational acting. (T) (F) 14. I have trouble changing my behavior to suit different people and different situations. (T) (F) 15. At a party, I let others keep the jokes and stories going. 41 (T) (F) 16. I feel a bit awkward in public and do not show up quite as well as I should. (T) (F) 17. I can look anyone in the eye and tell a lie with a straight face (if for a right end). (T) (F) 18. I may deceive people by being friendly when I really dislike them. 42 APPENDIX A: Pretesting Measures Strong Arguments Lately there has been much concern in the entertainment community to help those Americans and other Nationals that are hungry and homeless. Only months ago, Hands Across America took place to raise funds for the hungry and homeless in America. Later this fall, Home Aid will take place. Home Aid is a group of entertainers, such as movie and television stars, theatrical performers, and recording artists. Although many people view Home Aid as a good idea and deserves our support, I believe that Home Aid is a rather poor idea and deserves our nonsupport. Let me tell of some of the reasons why I feel this way. First, the concerts are too commercialized and cost too much to put on. The publicity and hype surrounding the concert is extensive. The promoters of the concert must pay advertising agencies for their services with the money raised for the needy. In addition, stage crews, managers, and directors must be paid. Further, there is a rental charge to use the stadium where the event takes place. Clearly, it appears that the costs of the production are too great. Secondly, after all the expenses have been paid, there's no real way of knowing that all of the remaining money actually gets to the needy. The money must be channeled into governmental and private agencies that have been set up to help the needy. However, because of the bureaucracy characterizing these agencies, embezzlement of these funds is accomplished rather easily. Lastly, and perhaps most importantly, people will think that the plight of the hungry and homeless are solved. At best, the funds that do get to the needy provide a temporary solution to a long term problem. We won't be able to cure the problem with concerts. Much more needs to be done to help the hungry and homeless. Because of these reasons, I urge you not to support Home Aid. 1. How valid do you think the arguments in the preceding paragraphs are? 1 2 3 4 5 6 7 not at all very valid valid 2. How true do you think the arguments in the preceding paragraphs are? 1 2 3 4 S 6 7 not at all very true true 43 3. How strong do you think the arguments in the preceding paragraphs are? 1 not at all strong very strong 44 APPENDIX A: Pretesting Measures Weak Arguments Lately there has been much concern in the entertainment community to help those Americans and other Nationals that are hungry and homeless. Only months ago, Hands Across America took place to raise funds for the hungry and homeless in America. Later this fall, Home Aid will take place. Home Aid is a group of entertainers, such as movie and television stars, theatrical performers, and recording artists. Although many people view Home Aid as a good idea and deserves our support, I believe that Home Aid is a rather poor idea and deserves our nonsupport. Let me tell of some of the reasons why I feel this way. First, by supporting Home Aid, you show the promoters of the event that you are not tired of these carbon-copy concerts. We have already experienced Live Aid, Band Aid, and Farm Aid--do we need to experience Home Aid too? By not supporting Home Aid, you will be showing the promoters and organizers that we want a fresh approach. These concerts seem all alike. Secondly, there is too much repetitive music in the concerts and not enough of a variety of performers. The promoters and organizers only invite top 40 bands to play. After you have listened to a few songs, you heard them all. We need a wide range of artists and types of music to show that everyone cares about the needy, not just a few. Lastly, the concert only allots its performers a given amount of time to perform. After watching and waiting for hours to see and hear your favorite artist, five to ten minutes does not seem long enough. I would rather go to a concert where my favorite performer is playing and make a donation there. Because of these reasons, I urge you not to support Home Aid. 1. How valid do you think the arguments in the preceding paragraphs are? 1 2 3 4 5 6 7 not at all very valid valid 2. How true do you think the arguments in the preceding paragraphs are? 1 2 3 4 5 6 7 not at all very true true 45 3. How strong do you think the arguments in the preceding paragraphs are? 1 not at all strong very strong APPENDIX B: Experimental Materials 46 APPENDIX B: Experimental Materials EXPERIMENTER INSTRUCTIONS 1) First, erase the blackboard and remove any papers or debris from the desks or tables in the room. 2) Take the desk or table that is in the front of the room so that you will be facing the subjects. 3) Place the testing materials in an orderly manner in front of you. 4) When the subjects arrive do not appear overly friendly or assertive. Conduct yourself in a professional manner. 5) Wait about 4 minutes before starting the experimental session in case anyone shows up late. Once you start a session, no one may be admitted. 6) Introduce yourself, say, "Hi, my name is ______ Is everyone here for an experiment called Reactions to Home Aid? What we are going to ask you to do is to read and react to a celebrity statement urging nonsupport of Home Aid. We also have another questionnaire that we would like you to complete, in which we are interested in how people make predictions." "Who needs a pencil?” (Pass out pencils). Now, give each subject a subject number, and say, "This number on this small piece of paper is your subject number. I would like you to put this number on each of your questionnaires, in this way we can match up all your questionnaires without identifying who you are--your responses will remain anonymous. Please do not write your name on any of the questionnaires." "What I would like you to do right now is to read and then sign the consent form.” (Pass out consent form). 7) When everyone has signed the consent form, say, ”A Detroit-based consulting firm, Webster, Webster, 8 Maxell has asked us to help them with some market research. They are interested in knowing if there is any student support for a charity event that would raise money for the hungry and homeless. The event would take place later this fall. They have already contacted a few celebrities, but some of the celebrities were not sure whether to participate in the fund raiser. In recruiting the celebrities, the marketing firm said, 'Lately, there has been much concern in the entertainment industry to help those Americans and other Nationals that are hungry and homeless. Only months ago, Hands Across America took place to raise funds for the hungry and homeless in 47 America. Later this fall, Home Aid will take place. Home Aid is a group of entertainers, such as, movie and television stars, theatrical performers, and recording artists. Could we count on your support by giving a short performance?‘ One of the celebrities contacted was ______ This was his reaction to the firm's request. Please take the next two minutes to read the statement." (Pass out the statement blank side up). Say, "OE, turn over the page and spend the next two minutes reading the statement." (Start timing subjects). 8) When two minutes are up, pass out the attitude toward Home Aid measure and collect the statement. 9) When everyone has completed the attitude measure, pass out the reaction to Home Aid measure. 10) When this is complete, pass out the thought listing measure. Say, "Please take the next seven minutes to list all the thoughts, ideas, and associations that you had about the statement as you read through the message. When seven minutes are up, I'll call time. You may begin." (Start keeping time). 11) Now say, "OK, time. Now go onto the next page of the questionnaire." 12) When everyone is finished, pass out the manipulation check measure. 13) When that questionnaire is complete, say, "Now I would like you to complete this." (Pass out the Detroit Tiger probability questionnaire). 14) When everyone has completed that measure, pass out the Martina Navratilova probability questionnaire. 15) When everyone has completed it, say, "Just one more thing before you go." (Pass out the recall measure). 16) Before dismissing the subjects say, "I do not know how the results will turn out, however, if you would like to receive some information about the study, please print your name and address on an envelop that I am passing around. Thanks for helping us out." 48 APPENDIX B: Experimental Materials Attitude Toward Home Aid SUBJECT NO _____ the number that you circle the For each of the following adjectives, best describes your opinion of Home Aid. please circle For example if 4, it means that you feel neutral. 1. Good 1 2 3 4 5 6 7 not at all very much so 2. Valuable 1 2 3 4 5 6 7 not at all very much so 3. Cruel 1 2 3 4 5 6 7 not at all very much so 4. Justified 1 ’ 2 3 4 5 6 7 not at all very much so 5. Needed 1 2 3 4 5 6 7 not at all very much so 6. Worthy 1 2 3 4 5 6 7 not at all very much so 49 APPENDIX B: Experimental Materials Reactions to Home Aid SUBJECT NO _____ For each of the following adjectives, please circle the number that best describes how you feel right now. For example, by circling the 4, it means that you feel neutral. 1. Pleasant 1 2 3 4 5 6 7 not at all very much so 2. Happy 1 2 3 4 5 6 7 not at all very much so 3. Awful 1 2 3 4 5 6 7 not at all very much so 4. Glad 1 2 3 4 5 6 7 not at all very much so 5. Helpful 1 2 3 4 5 6 7 not at all very much so 6. Caring 1 2 3 4 5 6 7 not at all very much so 50 7. Foolish 1 2 3 4 5 6 7 not at all very much so 8. Patriotic 1 2 3 4 5 6 7 not at all very much so 51 APPENDIX B: Experimental Materials Thought Listing SUBJECT NO _____ In the spaces provided below, please take the next seven minutes and list all of the thoughts, ideas, and associations you had about the statement as you read through the message. (Note: You need not fill up the entire page, just list as many as you can). 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 52 53 Now, I would like you to go back to the thoughts that you just listed and put a "+" next to all thoughts that you feel are in support of the writer's position, a "-” next to all thoughts that you feel do not support the writer's position (i.e., counterarguments), and a "0" next to all thoughts that you feel are neutral with respect to the writer's position. 54 APPENDIX B: Experimental Materials Manipulation Check and Additional Items SUBJECT NO _____ 1. I have heard of or know the writer of the statement. yes _____ no _____ 2. To what extent do you like the writer of the statement? 1 2 3 4 5 6 7 not at all very much 3. To what extent do you agree with what the writer has said? 1 2 3 4 5 6 7 strongly neutral strongly disagree agree 4. Have you ever heard of public fund raisers such as Band Aid, Farm Aid, or Hands Across America? yes _____ no _____ 5. How familiar would you say you are with such public fund raisers such as, Band Aid, Farm Aid, or Hands Across America? 1 2 3 4 5 6 7 not at all very familiar familiar 6. Have you ever participated in a public fund raiser such as Band Aid, Farm Aid, or Hand Across America by either volunteering your time or pledging support? YQS _____ no _____ 55 7. How important to you personally are the issues surrounding what the promoters of Band Aid, Farm Aid, or Hands Across America are trying to accomplish? 1 2 3 4 not at all important 6 7 very important 8. How important do you think are the goals of such fund raisers such as, Band Aid, Farm Aid, or Hands Across America? 1 2 3 4 not at all important 9. To what extent do you think the physically attractive? 1 2 3 4 not at all 10. To what extent do you think the appeal? 1 2 3 4 has none at all 11. To what extent do you think the good personality? 1 2 3 4 not at all good 12. To what extent do you think the personality? 1 2 3 4 not at all attractive personality writer writer writer writer of the of the of the has an 6 7 very important statement is 6 7 very much statement has sex 6 7 has very much statement possesses a 6 7 very good attractive 6 7 very much attractive personality 56 APPENDIX B: Experimental Materials Detroit Tigers Probability Questionnaire SUBJECT NO _____ Suppose the Detroit Tigers reach the 1987 World Series. Using the following scale, estimate the probability of each of the following events occurring. Please read all the events carefully before making your estimates. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% very very low high _____ Detroit will win the World Series. _____ Detroit will lose the first game of the World Series. _____ Detroit will win the first game of the World Series, but will lose the World Series. _____ Detroit will lose the first game of the world Series, but will win the World Series. How would you rate your knowledge of professional baseball? (circle one). well below below average above well above average average average average Who were last year's four division winners? Last year's most valuable player? Last year's rookie of the year? Have you ever played organized baseball? _____ yes _____ no If yes, for how long? 57 APPENDIX B: Experimental Materials Martina Navratilova Probability Questionnaire SUBJECT NO _____ Suppose Martina Navratilova reaches the 1987 Wimbledon finals. Using the following scale, estimate the probability of each of the following events occurring. Please read all the events carefully before making your estimates. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% very very low high _____ Martina will win the match. _____ Martina will lose the first set. _____ Martina will win the first set, but lose the match. _____ Martina will lose the first set, but win the match. How would you well below average rate your knowledge of professional tennis? above average below average average (circle one). well above average Who were last year's winners in the following tournaments? Men's Singles Women's Singles French Open Wimbledon Australian Open 0.8. Open 58 Have you ever played organized tennis? If yes, for how long? _____ yes _____ no 59 APPENDIX B: Experimental Materials Recall Measure SUBJECT NO _____ Using the spaces provided below, please recall as many of the arguments favoring your nonsupport of Home Aid as you can. (Note: You need not fill up all the spaces, just list as many as you can). 1. 10. LI ST OF REFERENCES 60 LIST OF REFERENCES Axsom, D., Yates, 8., 8 Chaiken, S. (1987). Audience response as a heuristic cue in persuasion. Journal of Personality and Social Psychology, 5;, 30-40. Borgida, E., 8 Howard-Pitney, B. (1983). Personal involvement and the robustness of personal salience effects. Journal of Personality and Social Psychology, 5;, 560-570. Brock, T. C. (1967). Communication discrepancy and intent to persuade as determinants of counterargument production. Journal of Experimental Social Psychology, 3, 296-309. Cacioppo, J. T., 8 Petty, R. E. (1984). The need for cognition: Relationship to attitudinal processes. In R. P. McGlynn, J. E. Maddux, C. D. Stoltenberg, and J. H. Harvey (Eds.), Social perception in clinical and counseling psychology. Lubbock: Texas Tech Press. Cacioppo, J. T., 8 Petty, R. E. (1985). Central and peripheral routes to persuasion: The role of message repetition. In L. F. Alwitt, 8 A. A. Mitchell (Eds.), Psychological processes and advertising effects. Hillsdale, NJ: Lawrence Erlbaum Associates. Cacioppo, J. T., Petty, R. E., 8 Sidera, J. (1982). The effects of a salient self-schema on the evaluation of proattitudinal editorials: Top-down versus bottom-up message processing. Journal of Experimental Social Psychology, 18, 324-338. Chaiken, S. (1980). Heuristic versus systematic information processing and the use of source versus message cues in persuasion. Journal of Personality and Social Psychology, 32, 752-766. Chaiken, S. (1982). The heuristic/systematic processing distinction ip persuasion. Presented at Society of Experimental Social Psychologists. Nashville, TN. Chaiken, S. (1986). The heuristic model. In Social Influence: The Ontario Symposium (Vol. 5). M. P. Zanna, J. M. Olson, 8 C. P. Herman (Eds). Hillsdale, NJ: Erlbaum. Craik, F. I. M., 8 Lockhart, R. S. (1972). Levels of processing: A framework for memory research. Journal of Verbal Learning and Verbal Behavior, 11, 671-684. 61 DeBono, X. G. (1986). On the systematic processing of persuasive message: Functional theories revisited. Unpublished manuscript, Union College. DeBono, K. G. (1987). Investigating the social adjustive and value expressive functions of attitudes: Implications for persuasion processes. Journal of Personality and Social Psychology, 5;, 279- 287. Erber, R., 8 Fiske, S. T. (1984). Outcome dependency and attention to inconsistent information. Journal of Personality and Social Psychology, 31, 709-726. Greenwald, A. G. (1968). Cognitive learning, cognitive responses to persuasion, and attitude change. In A. G. Greenwald, T. C. Brock, 8 T. M. Ostrom (Eds.), Psychological foundations of attitudes. New York: Academic Press. Harkness, A. R., DeBono, 8. G., 8 Borgida, E. (1985). Personal involvement and strategies for making contingency judgments: A stake in the dating game makes a difference. Journal of Personality and Social Psychology, 12, 22-32. Haugtvedt, 0., Petty, R. E., 8 Cacioppo, J. T. (1986). Need for cognition and the use of peripheral persuasion cues. Presented at the Annual Meeting of the Midwestern Psychological Association, Chicago, IL. Howard-Pitney, B., Borgida, E., 8 Omoto, A. M. (1986). Personal involvement: An examination of processing differences. Social Cognition, M, 39-57. Katz, 0. (1960). The functional approach to the study of attitudes. Public Opinion Quarterly, 2;, 163-204. Keppel, G. (1982). Design and Analysis; A researcher's handbook (2nd ed.). Englewood Cliffs, NJ: Prentice-Hall. Petty, R. E., 8 Cacioppo, J. T. (1981). Attitudes and persuasion: Classic and contemporary approaches. Dubuque, IA: William C. Brown. Petty, R. E., 8 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 Ps cholo , pg, 69-81. Petty, R. E., 8 Cacioppo, J. T. (1986). The elaboration likelihood model of persuasion. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 19). New York: Academic Press. Petty, R. E., Cacioppo, J. T., 8 Goldman, R. (1981). Personal involvement as a determinant of argument-based persuasion. Journal Personality and Social Psychology, 11, 847-855. 62 Petty, R. E., Wells, G. L., Heesacker, M., Brock, T. C., 8 Cacioppo, J. T. (1983). The effect of recipient posture on persuasion: A cognitive response analysis. Personality gpd Social Psychology Bulletin, 3, 209-222. Ratneshwar, 8., 8 Chaiken, S. (1986). When is the expert source more persuasion? A heuristic processing analysis. In T.A. Shimp, S. Sharma, et al., (Eds.), American Marketing Association Summer Marketing Educators' Conference Proceeding. Chicago: AMA. Sampson, E. E., 8 Insko, C. A. (1964). Cognitive consistency and conformity in the autokinetic situation. Journal of Personality and Social Psychology, 33, 184-192. Smith, M. B., Bruner, J. S., 8 White, R. J. (1956). Opinions and personality. New York: Wiley. Snyder, M. (1974). Self-monitoring of expressive behavior. Journal 93 Personality and Social Psychology, 39, 526-537. Snyder, M., 8 DeBono, 8. (1985). Appeals to image and claims about quality: Understanding the psychology of advertising. Journal of Personality and Social Psychology, 33, 586-597. Snyder, M., 8 Gangestad, S. (1986). On the nature of self-monitoring: Matters of assessment, matters of validity. Journal of Personality and Social Psychology, 31, 125-139. Wood, W., 8 Eagly, A. H. (1981). Stages in the analysis of persuasive messages: The role of causal attributions and message comprehension. Journal of Personality and Social Psychology, 32, 246-259. Wood, W., Kallgren, C. A., 8 Preisler, R. M. (1985). Access to attitude-relevant information in memory as a determinant of persuasion: The role of message attributes. Journal of Experimental Social Psychology, 13, 540-559. Yalch, R. F., 8 Elmore-Yalch, R. (1984). The effect of numbers on the route to persuasion. Journal of Consumer Research, 11, 522-527. 63 FOOTNOTES 1 Although both male and female subjects participated, no specific predictions were made regarding the combined effects of self-monitoring and gender, as none were suggested by previous research. Because no hypotheses were formulated regarding gender, the gender variable was collapsed in the subsequent analyses. 2 When the six items were collapsed into a single attitude measure, the item, cruel, was reverse scored so that all items were positively valenced. "IIIIIIIIII'IIIIIIIII