I ' 7‘ ; MAGNITUDE AND ATTITUDE CHANG ' 'TiiésTstor‘tiné neg-mm THE EFFECTS OF INFLUENCE FEEDBACKE“ AND NEED INFLUENCE ON THE RELATIONSHIP BETWEEN INCENTIVE??? MICHIGAN STATE UNIVERSITY . , _ if CHARLES R BERGER TTTTTTTTT . I; 1968 *Vw‘g,jh~~N T-m,,- ~ 0—169 L I B R A R Y Michigan State University IMIILIZIIQIBIIIINNIIuIIflILIIILIII I This is to certify that the thesis entitled THE EFFECTS OF INFLUENCE FEEDBACK AND NEED INFLUENCE ON THE RELATIONSHIP BETWEEN INCENTIVE MAGNITUDE AND ATTITUDE CHANGE. presented by Charles R. Berger has been accepted towards fulfillment of the requirements for Mdegree in_C9/’II’"H [I IC-q'IiC‘h Major professor Dam flfi‘m ‘4 /26 f I‘ , AU ‘0‘ 5:11 ABSTRACT THE EFFECTS OF INFLUENCE FEEDBACK AND NEED INFLUENCE ON THE RELATIONSHIP BETWEEN INCENTIVE MAGNITUDE AND ATTITUDE CHANGE by Charles R. Berger Dissonance theorists suggest that the less the justification (e.g. monetary incentives) for engaging in belief-discrepant behavior, the more individuals' attitudes will change in the belief-discrepant direction. Incentive theorists posit the Opposite relationship between incentive magnitude and self-persuasion. The present study sought to produce both dissonance and incentive effects, depending on persons' levels of need to influence (2 Influence). It was assumed that high 2_Influence persons would regard the opportunity to influence others as sufficient justification for engaging in belief-discrepant behavior; however, low 2 Influence persons should stress other loci of justification such as monetary incentives. Thus, it was predicted that at the level of simple commitment to engage in belief-discrepant behavior, there would be an inverse relationship between incentive magnitude and attitude change among persons with low E_Influence, while this relationship would not hold for high B_Influence persons, It was further assumed that when compared to low 2 Influence persons, high 2 Influence individuals would be more sensitive to cues indicative of successful influence of others. It was hypothesized that if a person with high 2_Influence took a belief-discrepant position on Charles R. Berger an issue and successfully persuaded another person to adopt that position, the high 2 Influence communicator would manifest greater self- persuasion than would a successful low 2_Influence communicator. Finally, in respect to overall self-persuasion, a third hypothesis predicted that high £_Influence communicators would display greater self-persuasion than would low 2 Influence communicators. Three hypotheses concerning message variables were also tested. Following dissonance theory, it was assumed that low 2_Influence persons encoding belief-discrepant messages under low incentive conditions would experience more congitive stress than would high 2 Influence encoders or low 2 Influence persons encoding under high incentive con- ditions. Based on the results of prior studies, it was predicted that low 2_Influence persons encoding under low incentive conditions would make more cross-outs and insertions, use fewer first person pronouns, and encode longer sentences than would persons in the other three conditions. Measures of 2 Influence and of attitudes toward college student draft deferments were initially obtained from undergraduate students during class time. Approximately three weeks later, subjects reported to the main experiment. Here, subjects were offered either $2.50 or 50 cents to write essays favoring the elimination of draft deferments for college students. Most subjects strongly opposed this position initially. Subjects were told that their essays would be read and reacted to by other students who also opposed the elimination of deferments. Charles R. Berger Subjects were paid before essay writing and their attitudes toward the draft deferment issue were assessed. After subjects wrote their essays, the essays were taken to another room where bogus information was put on each essay. Some subjects received comments and ratings which indicated successful influence, while other subjects received failure information. After subjects read the comments on their essays, attitudes toward the draft deferment issue were again measured. At the level of simple commitment, no support was found for the predicted inverse relationship between incentive magnitude and attitude change among low 2 Influence persons. Consistent with Hypothesis 2, high 2 Influence persons who received success feedback tended to manifest greater self-persuasion than did successful low 2 Influence persons. As predicted by Hypothesis 3, high 2 Influence persons displayed greater self-persuasion than did low 2 Influence persons. Finally, two of the three message hypotheses were partially supported: Although no significant effects were found for the sentences length measure, low 2 Influence persons in the low incentive condition and high 2 Influence persons in the high incentive condition used fewer first person pronouns and made more cross-outs and insertions in their belief-discrepant messages than did persons in the other two conditions. The results were discussed in light of relevant dissonance and incentive theory interpretations. Accepted by the faculty of the Department of Communication, College of Communication Arts, Michigan State University, in partial fulfillment of the require- ments for the Doctor of Philosophy degree. AZ/ICQDIRN7 6N? ;&VZQI§611, Director of Thesis Guidance Committee: , Chairman {imam I b Cum/- [/ZIMM/ZM 4’ fibmw THE EFFECTS OF INFLUENCE FEEDBACK AND NEED INFLUENCE ON THE RELATIONSHIP BETWEEN INCENTIVE MAGNITUDE AND ATTITUDE CHANGE B y \ a :é' a I Charles R. Berger A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Communication 1968 @570fl saw/grew am- I“. u e. n ACKNOLWEDGEMENTS The author would like to express his appreciation to Dr. Gerald Miller for his kind assistance in every phase of this study. Several of his ideas influenced the rationale for and the conduct of this study. Thanks also go to Dr. Erwin P. Bettinghuas, Dr. R. Vincent Farace, Dr. Hideya Kumata, and Dr. Frederick B. Waisanen who contributed their time and helpful suggestions. This study would not have been completed without the help of I several other persons. Ann Christiansen, Bette Blackburn, and Karen Pawlovich spent many hours coding 2_Influence protocols, while ElDean Bennett, Donald MacDonald, Roger Haney, and Edward Razinsky acted as experimenters. Adonna Kokx not only typed the final manu- script but also clarified many procedural details. Finally, my wife Joan and our two sons Douglas Paul and Daniel Riedel are to be thanked for thier patience and encouragement through- out the author's graduate studies. CHAPTER 1. INTRODUCTION: TABLE OF CONTENTS 2. METHODS AND PROCEDURES 0 3. RESULTS. 4. DISCUSSION . REFERENCES . o o o o o o 0 THE PROBLEM . 31 us 58 Table 10 Table 2. Table 3. Table u. Table 5 . Table 6. Table 7. Table 8. Table 9. Table 10. Table ll. Table 12. LIST OF TABLES Results of Message Analyses Performed in Six Self-Persuasion Studies . . . . . . . . . . . . Intercorrelations (r ) and Kendall's W for Three §_Influence Coders Sver Two Samples . . . . . . Means and Standard Deviations of Adequacy of Payment Ratings Across Eight Experimental Conditions . . . . . . . . . . . . . . . . . . Means and Standard Deviation of Perceived Adequacy Ratings Disregarding Feedback . . . . Means and Standard Deviations of Perceived Influence Measure Across Eight Conditions . . . Means and Standard Deviations of Perceived Persuasiveness Ratings Disregarding Monetary Incentive 0 O U 9 O O O 0 l 0 O O O O O O O U 0 Summary Analysis of Variance of Perceived Persuasiveness Means . . . . . . . . . . . . . Pre-test Simple Commitment Posttest Mean Attitude Change Scores and Amount of Change for Subjects in the Four Conditions . . . . . . Simple Commitment Posttest-Post-Feedback Posttest Mean Attitude Change Scores and Amount of Change for Subjects in the Eight Conditions . . . . . . . . . . . . . . . . . . Summary Analysis of Variance of Mean Simple Commitment-Post—Feedback Attitude Change Scores for Subjects in the Eight Conditions . . Mean Simple Commitment-Post-Feedback Attitude Change Scores Disregarding Monetary Incentive . Means and Standard Deviations of First Person Pronoun/Total Words Ratios Across Four Experi— mental Conditions . . . . . . . . . . . . . . . iv Page List of Tables (con't.) Table Table Table Table Summary Analysis of Variance of Mean First Person Pronoun/Total Words Ratios . . . . Means and Standard Deviations of Words/ Sentence Ratios Across Four Experimental Conditions . . . . . . . . . . . . . . . Means and Standard Deviations of Cross- Outs and Insertions/Total Words Ratios Across Four Experimental Conditions . . . Summary Analysis of Variance of Mean Error Ratios . . . . . . . . . . . . . . Page ALP] A‘— APPENDIX APPENDIX APPENDIX APPENDIX APPENDIX APPENDIX APPENDIX LIST OF APPENDICES TEST OF CREATIVE IMAGNIATION. OPINION PROFILE . n . . . . . TABLE Io - o o o a o o o o a TELEPHONE APPOINTMENT SCRIPT WKAR RADIO CAMPUS POLL . . . READER REACTION SHEET . . a o WMSB-TV STUDENT OPINION SURVEY vi a 72 73 74 81 82 Chapter 1 INTRODUCTION Es Problem Most research in communication and persuasion has focused on audience effects; that is, situations in which some source attempts to change audience attitudes. A second concern of persuasion researchers has been with individual differences related to the acceptance of social influence (Hovland and Janis, 1959). But until recently, relatively little attention has been paid to the effects of sources' encoding persuasiVe messages on their own subsequent attitudes and behavior. Within this general framework, the present study examines the more specific issue of attitudinal consequences of belief-discrepant message encoding behavior. Quite frequently, persons are induced or required to behave at variance with their private beliefs. For example, debaters frequently argue a side of an issue which is at odds with their own attitudes, and persons in formal organizations take public stands that are contrary to their private beliefs. What are the effects of such counterattitudinal behavior on a person's private beliefs? Moreover, what variables maximize or minimize attitude change in this self-persuasion situation? The present study examines several potentially relevant variables. In particular, the study deals with the effects of need to influence (2_Influence), success at persuasion, and monetary incentives on the magnitude of attitude change following the encoding of a belief—discrepant message. Moreover, consideration is given to the possible joint effects of 2_Influence and magnitude of monetary incentives on encoding perform- ance. It should be noted that this study represents the first attempt to employ E_Influence in a persuasion experiment. Furthermore, since the 2 Influence variable has been used in few prior studies, its use in the present study was considered exploratory. In this chapter, attention will be given to prior research relevant to the self- persuasion phenomenon. Also, a rationale and hypotheses will be presented for the present study. 22223 Research BEISelf-Persuasion A current, controversial problem in the area of communication and persuasion concerns the relationship between incentives for engaging in counterattitudinal behavior and attitude change following the counterattitudinal act. Early studies related to the self-persuasion problem (Janis and King, 1954; King and Janis, 1956) found that persons who actively engaged in belief-discrepant behavior showed greater attitude change in the advocated direction than did persons who passively lis- tened to a persuasive message. Furthermore, persons who improvised arguments in support of a belief—discrepant position changed their attitudes more than persons who merely read counterattitudinal materials. When discussing the results of the two studies cited above, Hovland, Janis, and Kelley (1953), argue that the superiority of improvised role-playing over passive listening in inducing opinion change is due to the tendency for persons in the role-playing situation to associate themselves with the belief-discrepant statements they make through implicit labeling responses. Hovland, Janis, and Kelly (1953) further assert: Thus the implicit labeling responses induced by improvisa- tion might sometimes be a key mediating factor that helps to explain why certain methods of eliciting overt verbal- ization facilitate opinion change whereas other methods fail to do so. (p. 236) This statement implies that when a person is induced to argue against his own beliefs, the more arguments he improvises supporting the belief- discrepant position and the more he associates himself with these arguments the greater his attitude change toward the new position. The position outlined above has been extended and elaborated in recent years by Janis and Gilmore (1965) and Elms and Janis (1965). These researchers are the main proponents of so-called incentive theory. In discussing incentive theory, Janis and Gilmore (1965) assert: According to this incentive theory, when a person accepts the task of improvising arguments in favor of a point of view at variance with his own personal convictions, he becomes temporarily motivated to think up all good positive arguments he can, and at the same time suppresses thoughts about the negative arguments which are supposedly irrele- vant to the assigned task. This 'biased scanning' in- creases the salience of the positive arguments and there- fore increases the chances of acceptance of the new attitude position. A gain in attitude change would not be expected, however, if resentment or other interfering affective reactions were aroused by ne atiVe incentives in the role- playing situation. (pp. 17-18) Janis and Gilmore assume that by increasing monetary incentives for en- gaging in counterattitudinal behavior, or by making other aspects of the social situation within which belief-discrepant behavior occurs more desirable, persons will be more highly motivated to improvise arguments in favor of the belief-discrepant position and more willing to identify themselves with the arguments they improvise. Thus, as incentives increase, opinion change toward the position advocated should also increase. In contrast to incentive theory, dissonance theory (Festinger, 1957), clearly predicts an inverse relationship between incentiVe magnitude and attitude change following a belief-discrepant act. Festinger (1957) states, "It also becomes apparent...that the magnitude of the reward or punishment, that is, the attractive- ness and desirability of the offered reward or the unpleasantness and undesirability of the threatened punishment, is an important determinant of the mag- nitude of dissonance which exists once compliance is exhibited. Too great a reward or punishment will result in only little dissonance. (p. 91)" The preceding statement implies that too much justification for en- gaging in belief-discrepant behavior will generate little dissonance and consequently result in little attitude change. There is a body of evidence which bears on both incentive theory and dissonance theory predictions. Festinger and Carlsmith (1959) offered subjects either 20 dollars or one dollar to tell an incoming subject (actually an accomplice) that a dull, boring task they had just completed was interesting and enjoyable. After engaging in this beliefudiscrepant communication behavior, subjects rated the task a second time. The results indicate that subjects in the one dollar condition rated the task significantly more enjoyable than did subjects in the 20 dollar condition. Since subjects in the one dollar condition supposedly received less justification, this result lends support to dissonance theory. In another study supportive of dissonance theory, Cohen (1962) offered subjects either 10 dollars, five dollars, one dollar, or 50 cents to write belief-discrepant essays. As in the Festinger and Carlsmith study, a significant inverse relationship was found between incentive magnitude and attitude change; i.e. subjects receiving small rewards manifested significantly more attitude change than subjects who received large rewards. Brehm and Cohen (1962) argue that under conditions of low justifi— cation, mere commitment to engage in a counterattitudinal behavior is sufficient to produce dissonance and subsequent attitude change. This position clearly runs counter to the incentive theory notion of impro- visation as a mediating factor in the production of opinion change. Hunt and Miller (1968) found support for Brehm and Cohen's position. In their study, subjects were instructed to encode belief-discrepant state- ments on the issue of disarmament. Before actually encoding the state— ments, the subjects” attitudes toward disarmament were assessed. Hunt and Miller found that subjects who scored high on Rokeach's dogmatism scale changed their attitudes significantly more than low dogmatic scorers. Moreover, subjects who were instructed to encode belief- congruent or irrelevant statements did not manifest significant attitude change. Thus, it appears that commitment to encode belief—discrepant messages may be sufficient to produce attitude change, at least among close-minded individuals. Rosenberg (1965) contends that subjects offered high rewards in the Festinger and Carlsmith study experienced evaluation apprehension. Evaluation apprehension is the tendency for participants in psychological experiments to think that the experimenter is evaluating their mental health. Rosenberg argues that under conditions such as being offered 20 dollars to lie, the subject may think that the experimenter is trying to determine whether he can be bribed. Thus, the subject perceives the experiment as a test of his honesty or his autonomy and resists atti- tude change under high reward conditions. Rosenberg further suggests that by separating the commitment and the reward phases of the self- persuasion experiment from the attitude measurement phase, evaluation apprehension can be reduced in the high reward condition. As a con- sequence, more attitude change should be manifested in'the high reward condition than in the low reward condition. To test his position, Rosenberg (1965) conducted a study in which be employed a two experiment guise to separate the belief-discrepant essay writing task from attitude measurement. Upon arriving at the experimenter's office, subjects were told that they would have to wait for the experiment to begin. The experimenter suggested that instead of waiting they could participate in another experiment being conducted by a "graduate student" in the Education Department. Subjects then reported to the "graduate student," who asked them to write counter- attitudinal essays. Before writing their essays, subjects were pro- mised either five dollars, one dollar, or 50 cents for their essay writing. After completing the task, subjects returned to the first experimenter. Subjects then filled out a questionnaire which contained items relevant to the belief-discrepant topic dealt with in their essays. Rosenberg's results strongly support the incentive theory position. Subjects in the five dollar condition changed their attitudes signifi- cantly more than subjects who received one dollar or 50 cents. While this evidence supports Rosenberg's prediction regarding the effects of evaluation apprehension, Aronson (l966) points to some problems in the Rosenberg experiment. Aronson argues that in the Cohen (1962) experiment, in which an inverse relationship was found between incentive magnitude and attitude change, the project was attributed to a presti- . gious sponsor. By contrast, Rosenberg's (1965) subjects were asked to participate in a little project being conducted by a "graduate student" in the Education Department. Aronson asserts that this induction lowered the prestige of the sponsor in the Rosenberg study. Aronson further suggests that subjects in the 50 cents condition did not take the essay writing task seriously because the 50 cent payment suggested that the experiment was a trival one being conducted in a low prestige department. In the case of the high reward subjects, the five dollar payment indicated that the experiment was important, even though it was being conducted in a low prestige department. Thus, the five dollar subjects tried harder and changed their attitudes more. It seems that Aronson‘s explanation for Rosenberg's results is very simi- lar to the incentive theory position which Aronson himself opposes. Studies performed by Janis and Gilmore (1965) and Elms and Janis (1965) shed some light on the sponsorship issue mentioned above, as well as the commitment only versus actual performance issue discussed earlier. Janis and Gilmore offered subjects either 20 dollars or one dollar to write counterattitudinal essays about disliked college courses. In addition, at each of the two levels of monetary incentive, some subjects were told that the experimenter was collecting information to be used by leading universities in the country (favorable sponsorship), while other subjects were informed that the essays would be used in an advertising campaign to be conducted by a book publisher (unfavorable sponsorship). Finally, some subjects actually wrote essays, while others did not. The findings indicate that subjects in the active writing-favorable sponsorship group changed their attitudes most, but the monetary reward variable had no significant effect on attitude change. Moreover, while not significant, attitude change among subjects who only committed themselves to write was consistent with dissonance expectations, i.e., the unfavorable sponsorship subjects showed more change. Elms and Janis (1965) conducted a study similar to that of Janis and Gilmore. Again, Elms and Janis manipulated favorability of sponsor- ship, overt versus no overt role-playing, and monetary incentive. Since Elms and Janis felt that the 20 dollar offer in the Janis and Gilmore study may have made subjects suspicious, thus interfering with impro~ visation and attitude change, subjects were offered a lower range of monetary incentives for their essay writing. The findings indicate that subjects who actually wrote essays for a favorable sponsor under condi— tions of high monetary reward showed the greatest attitude change. This pattern was not replicated in the no overt role—playing group. Both the Janis and Gilmore and the Elms and Janis studies support the incentive theory viewpoint regarding the relationship between justification and self—persuasion. In addition, the results cast some doubt on the generality of Brehm and Cohen's argument that mere commit- ment to engage in belief—discrepant behavior is sufficient to produce attitude change. It is, of course, seemingly difficult to reconcile the results of these two studies with the findings of the Festinger and Carlsmith (1959) and the Cohen (1962) studies cited earlier. In attempting such a reconcilation, Rosenberg (1966), argues that Festinger and Carlsmith subjects were set to deceive the "new subject." He asserts that this deception set will lead to an inverse relationship between incentive magnitude and attitude change. However, when a subject is asked to write a belief-discrepant essay, he must examine his position more closely. Rosenberg contends that this self—examining set results in a positive relationship between level of incentive and attitude change. But in developing this rationale, Rosenberg overlooks the findings obtained by Cohen (l962) which support an inVerse relation— ship between incentive magnitude and attitude change in an essay writing situation. A study conducted by Carlsmith, Collins, and Helmreich (1966) bears directly on the issue raised by Rosenberg. In one-half of their experiment, Carlsmith gt: a}: essentially replicated the Festinger and Carlsmith (1959) study. Subjects were given a dull task to perform and were offered either five dollars, one dollar and 50 cents, or 50 cents to tell an incoming subject (a confederate) that the experiment was interesting, exciting, and enjoyable. In the other half of the experiment, subjects performed the same dull task. But these subjects were then offered the same range of monetary incentives to write an essay about the positive nature of the experiment. Subjects in this condition were told that some ideas and points from their essays might be used to construct a new description of the task which would be read to succeeding subjects. However, they were assured that they would not be identified in any way. After the role-players had told the accomplice that the experiment was interesting and after the essay writers had finished their essays, ratings of the task were ob- tained in "another experiment" - a procedure employed to minimize the effects of Rosenberg's evaluation apprehension. 10 The results obtained by Carlsmith 333 3%. are relatively clear. Significant inverse relationships between incentive magnitude and ratings of interest and of fun were obtained in the face—to-face role- playing condition, while significant positive relationships between incentive level and ratings of interest and of fun were obtained in the essayuwriting condition. Thus, the role-playing findings support dissonance theory, while the essay findings support incentive theory predictions. Taken as a whole, the findings seem to support Rosenberg's notions of deception set and self-ekamining set as determinants of the relationship between incentive magnitude and attitude change. However, a study conducted by Linder, Cooper, and Jones (1967) casts doubt on Rosenberg's(l966) distinction between deception set and self—examining set. Linder 3:: 2%? reason that in the Cohen (1962) study discussed above, subjects had a relatively large amount of free- dom to choose 22£_£2 comply when asked to write belief-discrepant essays. In the Rosenberg study (l965) subjects had relatively little freedom to choose not to engage in the essay writing task. Linder £33 2%: argue that a person who has relatively greater freedom of choice to engage in belief-discrepant behavior will experience more dissonance than when he has little decision freedom. In their experiments Linder 2:: 2i. offered subjects either two dollars and 50 cents or 50 cents to write belief-discrepant essays. After the task was explained and the money offered, some subjects were explicitly advised that they could choose not to write the essay and could leave the experimental room. Other subjects were not explicitly given this option. Linder 253 2&3'3 results clearly support their ll hypotheses: In the high choice groups significant inverse relationships between incentive magnitude and attitude change were obtained, but this relationship was reversed in the low choice group. In contrast to the studies cited above, other researchers have found relationships between attitude change following a belief-discrepant act and positive or negative reinforcement given after performance of the behavior. Experiments performed by Scott (1957), Scott (1959), and Bostrom, Vlandis, and Rosenbaum (1961) support the notion that posi- tive reinforcement of belief-discrepant behavior leads to greater attitude change than does negative or no reinforcement. However, a study by Wallace (1966) suggests that rewarding different aspects of counterattitudinal performance may yield differential amounts of attitude change. In his study, Wallace had subjects assume a counterattitudinal position in a debate. At the conclusion of the debate some subjects received "audience ratings" indicating that their §£X$2.°f presentation was superior to other students, but that the content of their presentation was average. Other subjects were told that their style was average but that the content of their presentation was superior. A control group was told that they were average on both dimensions. Wallace found that subjects in the superior style- average content group manifested significantly more attitude change than did subjects in the other conditions. Several considerations emerge from this brief review of the self— persuasion literature. First, as Insko (1967) has pointed out, it appears that dissonance theory needs revision, at least in terms of predicting the relationship between incentive magnitude and attitude 12 change in the self-persuasion situation. Of course, this same need for revision is also relevant to incentive theory. Second, the results obtained by Carlsmith £53 a}? (1966) and Linder £5: 25: (1967) point to the necessity for searching out relevant variables that will delinate the circumstnaces under which incentive theory predictions will hold and the circumstances under which dissonance theory predictions will hold. It appears that neither theory is "correct," and that both seem to predict rather well in different situations. A final point concerns the attention that prior researchers have paid to subjects' message output. Both dissonance and incentive theories appear to have implications regarding encoding performance; however, relatively little effort has been directed at the investiga- tion of message differences among various treatment groups. Most of the attention given to such differences has involved the measurement of somewhat gross, subjective ratings of subjects' verbal output. Given the assumption that dissonance is tension producing, it seems reasonable to suggest that as tension increases, encoding per- formance should be impaired. Studies conducted by Stolz and Tannenbaum (1963), Miller (196“), and Combs (1968) generally support the proposition that negative audience feedback disrupts encoding performance. Greenberg (1960) found that subjects who encoded written messages under high cognitive stress took more encoding time, had lower rates of produc- tion (words/minute), made more grammatical, spelling and punctuation errors, produced less readable messages, relied more heavily on sub- stantive parts of speech (nouns, pronouns, and verbs), and relied less on descriptive parts of speech (adjectives and adverbs) than did subjects who encoded under low stress conditions. Bettinghaus and Preston (lean) 13 reported that subjects encoded belief—congruent statements more rapidly than belief-discrepant statements. This evidence suggests that sub- jects who encode belief-discrepant essays under conditions of low justification should display quite different encoding patterns than those who encode under conditions of high justification. This expectation is based on the assumption that subjects who encode belief- discrepant essays under low justification conditions experience more stress than subjects who encode under high justificatiOn conditions. Thus, according to dissonance theory, maximal decrement in encoding performance should be observed in the low justification (low reward) ‘group. Incentive theory predicts better encoding performance in the self—persuasion situation as incentives are increased, at least up to some optimal point. Table 1 contains the results of message analyses performed on the verbal output of subjects in several of the self-persuasion experi- ments cited earlier. Table 1. Results of Message Analyses Performed in Six Self-Persuasion Studies STUDY VARIABLES RESULTS Pestinger— 1. "Strength" of S's positive statements Carlsmith about the task. NSD 2. Rating of overall content NSD 3. Rating of persuasiveness of §_ NSD u. Amount of time S spent discussing task (not actual time, but rated) NSD Janis- 1. Arguments supporting position Sp in public Gilmore sponsor — $20.00 condition gave the largest num— her 2. Arguments against position NSD Table 1 (can't) 14 STUDY VARIABLES RESULTS Elms— 1. Quality of arguments Overt role- Janis players gave higher quality arguments in the favorable sponsorship con- dition. 2. Length of essay Higher paid §§ ‘ produced longer essays. Rosenberg 1. Length of essay Higher paid Sp produced longer essays 2. Basic persuasiveness of essay Higher paid Ss produced more persuasive essays Carlsmith l. Role-play performance 25, a}: a. Persuasiveness NSD b. Overall positiveness NSD c. Overall positiveness and conviction NSD d. Time spent on topic NSD e. Dissociation of self from content NSD 2. Accomplice's ratings of‘i a. Rated Ss on l.a., l.b., and l.c. NSD b. Apparent conflict $.50 Ss displayed more conflict c. Signs of discomfort NSD 3. Essay performance a. Emphasis used in making points NSD b. Elaboration of general theme NSD c. Overall quality and persuasiveness NSD d. Apparent effort NSD Linder 1. Experiment #1 et. al. a. Number of words per essay NSD "' "' b. Degree of organization NSD c. Intent of persuade NSD 2. Experiment #2 a. Persuasiveness NSD b. Degree of organization NSD c. Number of words NSD d. Extremity of position advocated NSD 15 It should be noted that many of the measures contained in Table 1 are based on judge‘s ratings, while few of the variables are counting measures such as those used by Greenberg (1960). Moreover, it is evident that in a majority of the experiments reported in Table 1, no significant differences among experimental conditions were found. Rationale and Hypotheses It was suggested above that one approach which would aid in clari- fying the relationship between incentive magnitude and attitude change following belief-discrepant behavior would be to seek variables that would yield interaction effects with the monetary incentive variable such that dissonance relationships would be obtained at one level of the variable and incentive effects at the other level. An alternative strategy would be to look for variables which yield neither dissonance nor incentive effects at one level of the variable and a dissonance or an incentive effect at the other level. The former strategy was chosen for the present study. The particular variable considered was n_Inf1uence. While attempting to devise a more reliable measure of Veroff's 2_Power (1955, 1958), Uleman (1965) found evidence for a new power~ related motive which he has labeled need to influence (2 Influence). Uleman's research stems from the work of McClelland on need for achieve- ment (McClelland, Atkinson, Clark, and Lowell, 1953; McClelland, 1961), and the research on needs for affiliation and power (Atkinson, 1958). Uleman's 2 Influence measure is based upon content analysis of Thematic Apperception Test (TAT) protocols. The 2_Inf1uence scoring system can be found in Uleman (1965). 16 Based on the results of several studies and correlations with extant personality inventories, Uleman (1965) characterizes persons with high 3 Influence as follows: those high in Uleman' s n Power do seem to enjoy the exercise of power for its own sake. They win more money in gambling games, even though the money is not ‘real' and there is no explicit incentive for winning...They are rated as more dominant by peers when dominance is defined in terms of self- confident mastery of others...power seems to be something to be used, rather than conserved defensively, and the goal is to see one s influ- ence on others, rather than to avoid being influ~ enced...Uleman' s n Power might be described as a need to influence. others for whatever purpose or goal. (p. 207) Uleman further points out that while Veroff's measure of n_Power apparently taps a person's need not to be influenced, high 2_ Influence persons seem to be oriented toward influencing others, rather than defending themselves against influence attempts. Given Uleman‘s evidence and his characterization of persons with high 2 Influence, two assumptions seem reasonable. Assumption 1: high 2 Influence individuals will be more sensitive to feedback regard- ing their success or failure at influencing another person or persons. Assumption 2: high 2 Influence persons will be less sensitive to the need for other justifications for engaging in belief-discrepant behavior, if they think they might influence another person or persons by producing belief-discrepant arguments. Furthermore, it is also reasonable to expect persons with high 2 Influence to put forth more effort than low 2 Influence persons when attempting to persuade others. Next, consider some implications of the above-mentioned assumptions when applied-to therehtionship between incentive magnitude and attitude change following belief-discrepant behavior. First, if l7 Assumption 2 is tenable, monetary incentives should have little, if any, effect on attitude change among persons with high n_Inf1uence; provided such persons perceive that the belief-discrepant material they encode may potentially influence other persons' attitudes and/or behavior. In contrast, low 2 Influence persons should view the opportunity to effect another's attitudes or behavior as relatively less important than the monetary incentives offered for engaging in belief-discrepant behavior. Thus, the monetary incentive variable should have maximal effects on attitude change, or self-persuasion, among low n_Inf1uence persons and minimal effects for high 2 Influence persons. Following Linder gt: 2&3'8 (1967) findings, an inverse relation- ship between incentive magnitude and attitude change would be expected among low n_Influence persons, if these persons felt that they had a choice not to engage in the belief—discrepant behavior. Moreover, if Brehm and Cohen's (1962) arguments are tenable, this inverse relation- ship should be observed at the level of simple commitment. These con- siderations lead to the following hypothesis: Hypothesis 1: At the level of simple commitment to engage in belief-discrepant behavior, there will be an inverse relationship between incentive magni— tude and attitude change among persons with low 2 Influence. This inverse relationship will not hold among persons with high 3 Influence. The relationships suggested in the above hypothesis will hold in the situation in which persons perceive that their belief-discrepant behavior might influence others' attitudes and/or behavior; that is, one would not expect to find differences in attitude change between high and low 2 Influence groups if the opportunity to exert influence were not part of the social situation within which the belief-discrepant behavior is performed. 18 It was suggested above that persons with high n_Influence will be more sensitive to feedback regarding their success or failure at influencing another person or persons. The Scott (1957), Scott (1959), Bostrom 23. ai. (1961), and Wallace (1966) studies all suggest that certain types of positive reinforcement for belief—discrepant behavior lead to greater attitude change than does negative or no reinforcement. In terms of n_Influence, positive reinforcement in the form of success- ful influence of another person should have more reinforcing value among high n_Inf1uence persons than among low n_Influence persons. Uleman (1965) has pointed out that in contrast to those with high n Power persons with high 3 Influence are not oriented toward defending themselves against other persons' influence attempts. Thus, if a person with high 2 Influence takes a belief-discrepant position on an issue and he successfully persuades another person to adopt that belief-discrepant position, the high 2 Influence communicator should manifest greater attitude change toward the position advocated than a low 2 Influence communicator who is also successful at persuasion. These relationships are stated in the following hypothesis: Hypothesis 2: After taking a belief-discrepant position, persons with high n_Inf1uence who success- fully persuade another to adopt that position will report greater attitude change in the belief-discrepant direction than successful low 2 Influence persuaders. However, there will be little difference in amount of attitude change between high and low E Influ- ence individuals who fail to persuade others to accept a belief-discrepant position. 19 The above hypothesis further implies that differences in attitude change between low n_Influence persons who fail and low 2 Influence persons who succeed in persuading should be small, but persons with high 2 Influence who succeed should display significantly greater attitude change than the highs who fail. A main effects hypothesis, which follows from Hypothesis 2, deals with differences in the amount of attitude change between high and low n_Inf1uence persons. Assuming that persons with high 2 Influ- ence put forth greater persuasive efforts, dissonance theory predicts greater attitude change among high n Influence individuals. Greater effort in supporting a counterattitudinal position produces more dissonance and leads to more dissonance reducing attitude change. Evidence supporting this position has been reported in a study by Zimbardo (1965). This line of reasoning leads to the following hypothesis: Hypothesis 3: After taking a belief-discrepant position, persons with high 3 Influence will report greater attitude change in the belief- discrepant than will persons with low 31 Influence. Next, consider some predictions concerning message differences among persons encoding belief-discrepant messages under the various conditions discussed above. Hypothesis 1 states that low 2 Influence individuals who encode under low justification conditions (low monetary incentive) should experience the greatest amount of disson- ance. This suggests that these persons will experience maximum tension, and that this tension will be reflected in their message output. The variables to be used to assess the level of this tension are taken from Greenberg (1960). 20 Greenberg (1960) found that journalism students who encoded messages under cognitive stress produced messages with more words per sentence, more cross-outs and insertions, and fewer first person pronouns per total words. These results are conSistent with his predictions; how- ever, Greenberg does point out that the first person pronoun per total words prediction might be reversed in populations other than journal; ism students, since journalism students are explicitly discouraged from using first person references in their writing. While this researcher believes that long sentences and high rates of cross~outs and insertions in messages may be indicants of stress, it is felt that the first person pronoun per total words ratio may function somewhat differently in the present study. Previously, it was argued that high 2_Influence persons should try harder to persuade others than individuals with low n_Influence. If this is the case, highs might therefore be more willing to identify themselves with their belief-discrepant arguments than lows. Thus, first person pronoun per total words ratios for the highs should be generally higher than for lows. However, persons with low n_Influence who receive high justifie cation for engaging in counterattitudinal behavior should also be more willing to identify themselves with their belief-discrepant arguments, while the lows who receive little justification should be less willing to engage in such identification. The above arguments suggest the following hypotheses for the following four experimental conditions: Low 2 Influence-High Justifi— cation, Low 5 Influence—Low Justification, High 2 Influence-High Justification, and High 2 Influence-Low Justification: 21 Hypothesis 4: Belief-discrepant messages encoded by persons with low 2 Influence under conditions of low justification will contain significantly fewer first person pronouns than belief—discrepant messages encoded in the other three conditions. Hypothesis 5: Belief-discrepant messages encoded by persons with low 2 Influence under conditions of low justification will contain significantly longer sentences than belief-discrepant messages encoded in the other three conditions. Hypothesis 6: Belief-discrepant messages encoded by persons with low n Influence under conditions of low justification will contain significantly more crOSSuouts and insertions than belief-discrepant messages encoded in the other three conditions. It should be emphasized that a given of the above three hypotheses is that the persons involved in encoding belief-discrepant communications perceive that they have the opportunity to influence other persons. It should also be noted that the feedback variable is not a part of these hypotheses, since its manipulation follows message encoding. Chapter 2 Methods and Procedures Overview To test the hypotheses presented in Chapter 1, this study was conducted in several steps: First, measures of attitudes and of 2 Influence were obtained during regular class time from students in communication and business writing courses. About three weeks later, most students tested in the classes were contacted by telephone and asked to participate in the second part of the study. This second I part was actually conducted about four weeks after the in-class measures were taken. In this phase of the study, the students reported to classrooms in a campus building and were then tested in small groups. After completion of the second part of the study all students were debriefed in their classes during regular class hours. Time I Procedures Before conducting the actual study, several pretests were carried out in order to identify an issue on which there was both general attitudinal agreement and considerable polarity. The issue which best met these two criteria was draft deferments for college students. Most students indicated that they strongly favored draft exemptions for college students. Based on the pretests two, seven category, Likert— type items dealing with student draft deferments were retained for further use. The procedures for collecting Time 1 attitude measures and the 2 Influence measure were as follows. The E was introduced by the regular 22 23 classroom instructor as a researcher from the Department of Communica- tion. After this brief introduction, E explained that §§ were to com- plete two tasks. The §_then distributed the booklets for the story writing {2 Influence measure) task. When all booklets were distributed, the instructions appearing on the front of the booklet were read aloud by the §_(Appendix A). The story writing task required about 20 minutes. At the conclusion of the story writing session, §§ were given a 25 item questionnaire, which contained the two draft deferment items for measuring Time 1 attitudes (Appendix B, Items, in and 23) above. Most §§ completed this questionnaire in about 10 minutes. After its completion, materials were collected and the E_thanked the class for their cooperation. The E then left the room. 3; Influence Measure The procedure used to measure 2 Influence involved the use of four of the Slide pictures used by Uleman (1965): l. The Army picture: Several military-looking men watching another soldier point to what looks like a map. 2. The Conference picture: Several men in business suits gathered around a conference table. 3. The Two Man picture: One man is seated, and the other is standing. The man who is standing is pointing to some papers the seated man is holding. H. The Newspaper picture: Man is seated reading a newspaper. 24 The slides were shown to §s in the above order. The Sp were instructed not to write their story while the slide was on the screen. After one minute of exposure, the slide was removed and the Sp were given four minutes to write a story about it. In order to guide the S'in his writing, four questions were asked on each writing sheet (Appendix A). This procedure was repeated for the other three pictures. Thus, each S wrote four stories. Three female graduate students were trained to code the stories for n_Influence. A brief description of the coding categories taken from Uleman (1966) appears in Appendix C. The score range for a particular story is zero through nine, where zero represents no influence imagery and nine represents maximum 2 Influence. Thus, when scores for the four stories are summed, an individuals's total n_Influence score may range from zero to 36. The results of inter- coder reliability checks on samples of the stories are presented in Chapter 3. Time II Design and Procedures One criterion was used to select Sp for the experimental phase of the study. Only §s who favored draft deferments for college students were used. Since two seven point items were used to assess attitudes toward draft deferments for college students, an individual's score could range from two to 19: a score of two represented strong opposi- tion to the elimination of draft deferments, while a score of 1a indi- cated strong endorsement of the idea of doing away with deferments. Any person who scored eight or more on the scale was not used in the experi- mental phase. Applying this criterion to the initial subject pool left 225 useable Sp. 25 The n_Influence scores for the 225 §s were split at the median. Sp in the high and the low 3 Influence groups were then randomly assigned to one of the following four treatment combinations: High Incentive ($2.50)-Success Feedback, High Incentive-Failure Feedback, Low Incentive (50 cents)-Success Feedback, or Low Incentive-Failure Feedback. Five paid undergraduate students contacted the 225 §s by telephone. The telephoners read a prepared script (Appendix D), which stated that the caller was from the production staff of WMSB television. The telephoners told the prospective Sp that WMSB, in collaboration with WKAR radio and the Television-Radio Department, would be producing a series of shows dealing with students' problems. Furthermore, the §§ were told that they had been selected randomly to give their opinions and ideas on these problems. Sp were asked to report to a specific room in a campus building, and were given a choice among several nights and times to participate in the project. Tele- phoners were directed not to mention the terms experiment or §££dy_in connection with the project. Sp were told that they would receive payment for their time, but they were not told how much money they would receive. Also, Es were not informed about the particular issue they would be dealing with or the exact nature of the task they would be asked to perform. The telephoners were able to contact 190 of the original 225 Sp. Only 10 of the 190 §s contacted indicated that they would be unable to participate in the project. 0f the 180 who agreed to participate, 134 actually came to the experiment. 26 Upon arriving at the designated room, §s were directed to report to one of two rooms. §§ assigned to the High Incentive condition were sent to one room, while Low Incentive condition E? were sent to the other. Two Es were used in each room. The first §_to enter each room (El) read the following instructions: "I'm from the production staff of WKAR radio. We would like to thank you for coming to our project. I am assisting the project director, and he will be here in a few minutes to explain what we will be doing for the next 30 minutes or so. While we are waiting for the project director to come, we would appreciate it if you would fill out this short campus poll being conducted by WKAR radio. Please read the instruc- tions carefully before filling out the poll." After these instructions were read, E1 distributed the WKAR ques— tionnaire (Appendix E). This questionnaire contained the two critical items from the Time 1 instrument as well as numerous filler items. A few seconds after the Es began filling out their questionnaires, the project director (E2) entered the room. E1 directed the Sp to stop writing and to listen to E2'8 instructions. El then introduced E2 who read the following instructions: "First, I would like to express my thanks for your coming here this evening. I hope that you will find this project interesting as well as enjoyable. Let me tell you about the first part of the project. WMSB— TV is producing a series of shows about problems students face. One problem we have chosen to deal with is the military draft. We are particularly interested in the issue of draft deferments for college students. In order to gather material on this issue, we are paying selected students to write scripts in which they present arguments for and against draft deferments for college students. We are asking you to write scripts which contain onl arguments advocating the elimination of draft deferments for college students; that is, you are to argue that college student draft deferments should be done away with. We are paying you each ($2.50 or 50 cents) to write your scripts. 27 It is important that your scripts be as persuasive as possible, since we are going to test your arguments on a group of students down the hall. This is why some of you were assigned to this room and other students were assigned to other rooms. After you finish writing your scripts, they will be taken to the reading group down the hall. Each script will be read by a few students. They will react to your arguments and indicate whether your script persuaded them that college students should not receive draft deferments. Most of the students in the reading group favor the system of draft exemptions for college students, so it is your job to persuade them to accept the opposite view. After your scripts have been read by the reading group, they will be re- turned to you with the readers' reactions. Now that you know the nature of the project, if any of you do not wish to participate, you may leave at this time." After the instructions were read, Sp were paid either $2.50 or 50 cents, depending on the condition to which they were assigned. After payment, the Sp finished filling out the WKAR questionnaire. Since Sp were interrupted before they filled out the items dealing with draft deferments, responses to these crucial items in the WKAR questionnaire were compared with responses made to the same items at Time 1, in order to assess attitude change as related to incentive magnitude and level of 2 Influence. It should be kept in mind that at the time of this measurement, no essays had been written by any Sp. After the Sp finished the WKAR questionnaire, they were handed three 8 1/2 x 11 sheets of paper on which to write their essays. Sp were giVen about 10 minutes to complete the task. During this time El and S2 worked together handing out and collecting materials. When the 10 minutes were up, the essays were collected and 24 took the essays to another room. After E1 left the room with the essays, E2 read the following instructions: 28 "While we are waiting for your scripts to come back, we are going to tape record some free discussion among you. The program board of WMBS-TV is interested in improving the quality of so-called educational television. Perhaps some of you are aware of the fact the educational tele- vision is a relatively unpopular concept in this country. Our programming board would like to know why, so we would like you to give us your ideas about why educational tele— vision programs are not as popular as regular commercial television programs. I'll give you a few minutes to think about some possible reasons. Then we will turn the tape recorder on. This tape will be played back to the WMSB-TV program board. Of course, it will not be played on the air. Please don't hesitate to give us your ideas even if you have had little experience with educational TV. All ideas are welcome. Remember, we want to know why educa- tional TV programs are less popular than regular commer- cial programs. Now, take a few minutes to think about the issue. After reading these instructions, 52 waited for about one minute and then turned on the tape recorder. £9 was instructed to keep the group discussion on the topic and not to allow the group to wander into issues related to draft deferments. While the discussions were in progress in the experimental rooms, the two E45 and two graduate students wrote bogus comments on the essays. The following four comments were written on the essays of those Sp in the Success Feedback condition: 1. Your agruments were very strong, and they really changed my mind. 2. Your script sure sold me, maybe college draft deferments should be eliminated. 3. You convinced me. a. You changed my mind. You made very good points. Sp in the Failure Feedback condition receiVed the following four comments on their essays: 1. Your points were very weak. Your script didn't change my mind a bit. 2. Your arguments were poor. I'm not convinced. 3. You must be kidding. Your script is bad. a. You didn't convince me at all. 29 Care was taken to use different pencils and pens, and different writing styles for each of the comments. In addition to written comments a set of bogus rating scales (See Appendix F) was also given to each S3 Each supposed reader of the essay indicated his attitude toward the elimination of draft defer- ments for college students before and after reading the essay. The before and after ratings were recorded on two bipolar adjective scales (good-bad). Since there were four sets of before-after scales on each sheet, Sp were led to believe that four persons had read and rated their essays. Sp in the Success Feedback condition received sets of scales indicating marked before-after change in the belief-discrepant direction. Failure Feedback condition Sp received sets of scales with patterns of markings indicating little or no change. Both the success and failure marking patterns were the same for Sp in the respective treatments, but the ordering of the marking differences varied among Sp in each of the two conditions. After the comments were written on the essays, the two E15 returned to their respective rooms. When Ed entered the room, E2 said: "Here are the scripts you wrote along with the comments of the students who read them. It looks as if some of you really convinced your readers, while some of you didn't. I'll hand back the scripts so that you can see how well you did." The essays were returned, and the Sp were given about two minutes to read the comments and to look at the rating scales. The essays were then collected and E2 read the following instructions: 30 "For the final task this evening, we would like to haVe your opinions on some issues we are planning to use in some programs. This is a short questionnaire, which will only take a few minutes to fill out. Please read the instructions on the cover carefully before you begin. Please answer all of the questions and put your name and student number in the spaces provided." This final questionnaire also contained the same two crucial items on draft deferments and was attributed to WMSB-TV (Appendix G). Sp responses to these two items were used to assess post-encoding attitude change following success or failure feedback. After filling out this questionnaire, Sp were thanked for their help. Sp in the 50 cent condition were given an additional two dollars for their participation; hence, all Sp were eventually paid $2.50 for their participation. Sp were asked not to tell their friends about the project, since advanced knowledge of the task might influence the kinds of essays written. The entire experimental pro- cedure took about us minutes to complete. Throughout the experimental phase, Eq'EQ pairs were alternated between incentive conditions in order to distribute possible biasing effects equally across the two treatments. l Chapter 3 Results 2 Influence Intercoder Reliability Two separate reliability checks were run on the three females who coded the 225 subjectS' n Influence protocols. During the coding procedure, the three coders scored the same HO subjects (160 stories). The total score rankings of these no subjects were compared among the three coders. Thus, approximately 17 percent of the 225 subjects' 2 Influence protocols were scored by all three coders. Generally, the intercoder reliabilities, expressed in terms of Spearman's rank correlation coefficient, were rather low. The inter- correlations among the three coders for each set of 20 subjects and the overall correlations across both sets are summarized in Table 2. Kendall's W (Siegel, 1956) was also used to assess the amount of agree- ment among all three coders. Table 2. Intercorrelations (rs) and Kendall's W for Three 2 Influence Coders Over Two Samples Sample 1 Sample 2 Total (2p20) (2320) (N=40) rS (1-2) ,52* .33 ,n7ssen rs (2-3) .79** .39* .62**** ps (1-3) .55** .3g* ‘55esse Kendall's W .73** .5 * .69*** *p (005 **p <.Ol seep (.QOL sweep (00005 31 32 In addition to the correlational data presented above, dis- agreements between pairs of coders were also analyzed for the total sample (N=u0). When coders disagreed to the extent that one coder placed a given subject in the high n_Influence group, while the other coder placed the same subject in the low n_Influence group (high-low split based on the grand median), the direction of the disagreement was noted. Among the three pairs of coders there was no systematic bias in the direction of these disagreements as determined by the Binomial test (Siegel, 1956); that is, when disagreements involving a median crosscover occurred, the two possible directions of disagree— ment between coder pairs were approximately equally likely. While Table 2 indicates that the three coders were not making random judgments, the magnitude of the correlations are below the generally accepted level of .80 to .85. The correlations for Sample 2 are especially low. However, low reliability in the 2_Influence measure lowers the probability of finding significant differences be- tween 3 Influence groupS. Thus, if significant differences between 2 Influence groups are observed, it can be concluded that level of‘n Influence is a powerful predictor. Manipulation Checks The last item on the final questionnaire was an open-minded item which asked each subject to indicate what he thought the project was about. Of the 134 subjects who participated in the experimental phase of the study, five gave responses which strongly indicated knowledge of the real purpose of the project. These subjects were eliminated from the analysis. Thus, data for 129 subjects ~ 92 males and 37 females - were analyzed. 33 In addition to the preceding question, subjects rated the adequacy of payment for the essay writing task. Adequacy of payment was measured on a four point scale ranging from Very Adequate (u) to Very Inadequate (1). Since adequacy of payment measures were obtained after the success- failure manipulation, the effects of this manipulation on the adequacy of payment ratings must be conSidered. The means and standard deviations of the adequacy of payment ratings across the eight experimental conditions are shown in Table 3. The data reflect a relative lack of variability in the adequacy ratings. Thus, even though mean differences are small, some of the differences are significant. In addition, it can be noted that no . group rated the payment below the theoretic median of the adequacy scale (2.5). Table 3. Means and Standard Deviations of Adequacy of Payment Ratings Across Eight Experimental Conditions* Success Feedback Failure Feedback 3 Influence High Low High Low Incentive Incentive Incentive Incentive High 3? = 3.58 3.55 3.80 3.25 S = .79 .58 .60 .69 p. = 19 17 10 20 Low 32' = n.00 3.05 l$.00 3.11). S = .00 .59 .00 .83 a = 11 18 17 in *Three of the 129 subjects did not complete this 1 item. 34 In terms of Hypothesis 1, little difference in perceived adequacy of payment should be expected among high 2_Influence subjects, but a large difference in perceived adequacy should be observed among the lows; that is, an interaction between 2_Influence and monetary incene tives. Of course, subjects who were paid $2.50 would be expected to rate their payment as more adequate than would subjects who received the 50 cent payment. Since there was no variability in two of the eight treatment groups, it was inappropriate to perform an overall analysis of variance of the eight mean ratings of payment adequacy. However, a tftest of the differences between the mean adequacy ratings for High versus Low Monetary Incentive conditions was calculated. The mean adequacy rating for the High Monetary Incentive condition was 3.83, while the mean for the Low Monetary Incentive group was 3.27. The difference between the two means is significant beyond the .0005 leVel, one-tailed test. Thus, subjects who received $2.50 perceived the payment as more adequate than did subjects who received 50 cents. It should be noted, however, that the Low Monetary Incentive group did not rate their payment below the theoretic midpoint of the adequacy scale. While it was not possible to directly test the interaction due to lack of within cell variability, the data found in Table u shed some light on this issue. The difference between the means for the High E'Influence-High Monetary Incentive and the High 2 Influence—Low Monetary Incentive groups was not significant (£.= 1.2”, df = 61+)° While no test of the difference between the means for the two Low 31 Influence groups was possible, it is apparent that Low 2 Influence 35 subjects who received $2.50 for their essay writing perceived their payment as more adequate than did the Low 2 Influence subjects who received 50 cents. These findings are consistent with the assumption that magnitude of monetary incentive is more important to low IL Influence than to high 2_Influence subjects, giVen that both types perceive that they have an opportunity to exert social influence. Generally, then, the results indicate that the monetary incentive manipulation was successful, and that the pre-conditions for attitude change predicted by Hypothesis 1 were met. Table 4. Means and Standard Deviations of Perceived Adequacy Ratings Disregarding Feedback High Incentive Low Incentive 3 Influence _ X = 3.65 3.u3 High S = .72 .69 g = 29 37 Y = u.00 3.09 Low 8 = .00 .68 2.: 28 32 A second manipulation check was also included in the final questionnaire. All subjects were asked to indicate on a four point scale the extent to which they felt that they had influenced the script readers' opinions on the draft deferment issue. This check served to test the effectiveness of the success-failure manipulation. A score of four on the item indicated very high perceived influence, while a score of one indicated very low perceived influence. 36 Table 5 summarizes the means and standard deviations for the perceived influence ratings for the eight conditions. As in the case of the perceived adequacy of payment meaSure, it was not possible to calculate an overall analysis of variance of the eight means, due to the absence of variability in one of the cells. It was possible, however, to carry out an analysis of variance using 2 Influence and Feedback as independent variableS, ignoring the monetary incentive classification. This analysis seemed appropriate, since the major purpose was to assess the effectiveness of the feedback manipulation. Table 5. Means and Standard Deviations of Perceived Influence Measure Across Eight Conditions* Success Feedback Failure Feedback n Influence - High Low High Low Incentive Incentive Incentive Incentive High Y = 3.58 3.32 1.00 1.20 S = .59 .56 .00 .51 2 = 19 19 10 2o 76 = 3.36 2.61 1.12 1.1L; Low S = .64 .95 .32 .35 a = 11 18 17 1L1 *One of the 129 subjects did not complete this item. The means and standard deviations of the perceived persuasiveness measures ignoring the monetary incentive classification are found in Table 6. These measures were analyzed uSing an approximation procedure for unequal 2:3 suggested by Winer (1962). 37 Table 6. Means and Standard Deviations of Perceived Persuasiveness Ratings Disregarding Monetary Incentive 2_Influence Success Feedback Failure Feedback f: 3.94 1.33 High 8 = .60 .38 2_= 38 30 35: 2.90 1.12 Low S = .76 .34 2_= 29 31 The results of this analysis, summarized in Table 7. clearly indicate that the feedback manipulation generated perceptions of successful influence in the Success group and of failure in the Failure group. Moreover, consider the significant interaction between 2_Influence and Feedback. It will be recalled that Hypothesis 2 states that high 2_Influence persons who take a belief-discrepant stand and who persuade another to adopt that position will manifest greater attitude change toward the belief-discrepant position than will low 2_Influence persons who succeed at the same task. Moreover, Hypothesis 2 predicts that high and low E_Influence individuals who fail will display about the same amount of change, but that this change will be minor compared with those who succeed. 38 Table 7. Summary Analysis of Variance of Perceived Persuasiveness Means Source of Variation ss df MS F 2_Influence (A) 0.0770 1 0.0770 6.47* Feedback (B) 4.16414 1 4.16““ 3119.95“: A x B 0.071%8 1 0.07148 6.29": Error 12% 0.0119 *p <.025 *kp <.001 The pattern of the mean perceived persuasiveness ratings found in Table 6 and the significant interaction obtained for these means support the assertion that the preconditions for the pattern of attitude change predicted in Hypothesis 2 were met. Using the error term from the analysis of variance of perceived persuasiveness means (Table 7), multiple trtests were calculated for the perceived persuasiveness means presented in Table 6. As would be expected, the difference between the means for the high and low 3 Influence subjects who failed was not significant (t_= .027, df = 59). The difference between the Success and the Failure means for the High B_Inf1uence and the Low 2_ Influence groups was highly significant, confirming the effectiveness of the feedback manipulation. The most important test involves compar- ison of the mean perceived persuasiveness rating of the High n_Inf1uence- Success group with the mean rating of the Low n_Influence-Success group. This comparison yielded a significant difference (£_= 3.63; df = 65; p_ <.0005, one-tailed). Thus, the feedback manipulation was not only successful in creating differential perceptions of success or failure to influence, but in addition, the effect was not equally distributed across the two 3 Influence groups. High 2_Influence 39 subjects who received success feedback felt that they had persuaded their essay readers more than did low E_Influence subjects who received success feedback. However, highs and lows who failed had similar perceptions of the persuasive impact of their essays. Taken as a whole, the results presented in this section indicate that the monetary incentive and the feedback manipulations were effective. In addition, it appears that the preconditions were established for the attitude changes predicted in Hypotheses l and 2. In the following sections the results for each hypothesis are considered. Hypothesis 1; At the level of simple commitment to engage in belief- discrepant behavior, there will be an inverse relation- ship between incentive magnitude and attitude change among persons withlow n Influence. This inverse rela- tionship will not hold among persons with high 2 Influ- ence. Since Hypothesis 1 is concerned with attitude change at the simple commitment level, the measure to assess this change was taken before the feedback variable was manipulated. Table 8 summarizes the amount of pretest-simple commitmentposttest change for subjects in the four conditions. It can be seen that little change occurred in any of the groups. Correlated Ertests for each of the four conditions revealed no significant changes. In addition, an analysis of variance on the mean change scores for the four groups yielded no significant diff— erences. Thus, the data provide no support for Hypothesis 1. Table 8. Pretest-Simple CommitmentPosttest Mean Attitude Change Scores and Amount of Change for Subjects in the Four Conditions Condition Pretest Simple Commitment Change High ‘1; Influence High Incentive (2f29) 3.97 9.28 +0.31 Low Incentive (£339) 3.12 3.17 +0.05 Low 2. Influence High Incentive (2f28) 3.25 3.75 +0.50 Low Incentive (2f33) 3.36 3.88 +0.52 9O Hypothesis 2; After taking a belief-discrepant position, persons with high 3 Influence who successfully persuade another to adopt that position will report greater attitude change in the belief-discrepant direction than successful low 2 Influence persuaders. However, there will be little difference in amount of attitude change between high and low n Influence individuals who fail to persuade others to accept a belief-discrepant position. Hypothesis 2 predicts maximum attitude change, or self-persuasion, among high 2 Influence persons who are successful in influencing others. Although the monetary incentive variable is not specifically involved in this prediction, it might somehow interact with the feedback variable; thus, monetary incentive was also included in the analysis for Hypothesis 2. Table 9 contains the simple commitmentposttestpost—feedback posttestattitude change scores for subjects in the eight conditions. It can be seen that a relatively large amount of attitude change occurred in both incentive conditions of the High p_Influence— Success group. This pattern of self-persuasion was not observed in the Low 2 Influence-Success group. Here, High Incentive subjects manifested significant attitude change, while Low Incentive subjects did not. Table 9. Simple CommitmentPosttest-Post-FeedbackPosttest Mean Attitude Change Scores and Amount of Change for Subjects in the Eight Conditions Simple Post- Condition Commitment Feedback Change High n Influence SucEess High Incentive (n=19) 9.32 6.32 +2.00*** Low Incentive (3:19) 3 .42 5 .05 +l.63°’~'=‘=* Failure High Incentive (pflO) 9.20 5.00 +0.80* Low Incentive (pf20) 2.95 3.95 +0.50 91 Table 9 (con't) Simple Post- Condition Commitment Feedback Change Low 3 Influence Success High Incentive (n=ll) 3.81 9.81 +1.00** Low Incentive (3219) 9.92 9.63 +0.21 Failure '- High Incentive (n=17) 3.71 9.52 +0.81* Low Incentive (2919) 3.19 3.35 +0.21 *2_ <.05 (one-tailed) sag <.01 (one-tailed) 9.9.2 <.005 (one-tailed) While the mean change scores presented in Table 9 provide some confirmation for Hypothesis 2, the major test of the hypothesis required a test of the relative change among the eight conditions.- Thus, the simple commitment-post-feedback change scores for subjects in the eight conditions were analyzed, again using the approximation procedure recommended by Winer. Table 10. Summary Analysis of Variance of Mean Simple Commitment-Post—Feedback Attitude Change Scores for Subjects in the Eight Conditions Source ofpygriation ss df MS F n Influence (A) 0.9000 1 0.9000 9.01999 Ihcentive (B) 0.5392 1 0.5392 2.38* Feedback (C) 0.7839 1 0.7839 3.9999 A X B 0.0666 1 0.0666 0.30 A X C 0.5825 1 0.5825 2.60* B X C 0.0077 1 0.0077 0.03 A X B X C 0.0015 1 0.0015 0.006 Error 121 0.2295 '« p <. 20 in?- < .10 LJ... 551.02 (005 92 This analysis (Table 10) indicated that only the main effect for 2 Influence was significant at the .05 level. The predicted inter- action between p_Influence and Feedback was not significant; however, as Table 11 reveals, the differences predicted by Hypothesis 2 are in the right direction. High p_Influence subjects who were successful at influencing their readers changed their attitudes more than success- ful low 2 Influence subjects. Moreover, it is interesting to note that this pattern of results is consistent with the findings concerning the perceived persuasiveness ratings (Table 6 and 7). Thus, success- ful high 2 Influence subjects not only saw themselves as more influen- tial than their low 2 Influence counterparts, they also changed their attitudes more; that is, demonstrated greater self—persuasion than successful low 2 Influence subjects. Table 11. Mean Simple Commitment‘Post-Feedback Attitude Change Scores Disregarding Monetary Incentive 2 Influence Success Feedback Failure Feedback High 7? = +1.83 +0.60 S = 2.32 1.86 p_= 38 30 Low 7 = +0.50 +0.59 S = 0.99 1.76 p_= 28 31 The preceding results provide some support for Hypothesis 2, but it is clear that the data represent only a trend in the predicted direction. The findings presented in Table 9 are encouraging, but the crucial test of relative change presented in Table 10 is somewhat 93 disappointing. Finally, examination of Table 11 reveals that the low 2 Influence subjects who succeeded changed their attitudes about as much as the low p_Inf1uence and high 2 Influence subjects who failed. Clearly, these data suggest the need for further experimentation re- garding the relationship between p_Influence, feedback regarding the outcomes of influence attempts, and attitude change following a belief-discrepant act. Hypothesis 3} After taking a belief-discrepant position, persons with high n Influence will report greater attitude change in the belief-discrepant direction than will persons with low 2. Influence . The results presented in Table 10 are directly relevant to the test of Hypothesis 3. The mean attitude change for the high 2 Influ— ence subjects was +1.27, while the mean change for the low 2 Influence subjects was +0.53. As Table 10 indicates, this differences is significant. Thus, the data suPport Hypothesis 3. Hypothesis 9; Belief-discrepant messages encoded by persons with low p_Inf1uence under conditions of low justification will contain significantly fewer first person pronouns than belief-discrepant messages encoded in the other three conditions. Hypothesis 9 is the first hypothesis concerned with the effects of 5 Influence and monetary incentives on belief-discrepant encoding behavior. None of these hypotheses involve the feedback variable, since feedback was not manipulated until after the essays had been written. Table 12 contains the mean first person pronoun/total words ratios and the standard deviations of these ratios for subjects in the four experimental conditions. 99 Table 12. Means and Standard Deviations of First Person Pronoun/ Total Words Ratios Across Four Experimental Conditions 2 Influence High Incentive Low Incentive High Y = 0.0131 0.0210 S = 0.0191 0.0173 g = 29 39 Low 2 = 0.0202 0.0198 S = 0.0200 0.0100 n = 28 33 Inspection of these means reveals that the pattern predicted in ‘ Hypothesis 9 was not obtained. As predicted, the means for the High 2 Influence-Low Incentive and the Low p_Influence-High Incentive groups are greater than the mean for the Low 3 Influence-Low Incentive group. However, the mean for the Low p_Inf1uence-Low Incentive group is higher than the mean for the High p_Influence-High Incentive group, a difference that runs counter to Hypothesis 9. Analysis of variance of the first person pronoun/total words ratios (Table 13) revealed a significant interaction between 2 Influence and Monetary Incentive. These results suggest the question of why subjects with high p Influence who received high incentives did not use a large number of first person pronouns. Consideration of the results obtained for Hypotheses 5 and 6 may shed further light on this question, for Greenberg (1960) has suggested that the results observed for one index of cognitive stress may or may not be consistent with other stress indices, depending on the situation in which encoding takes place. 95 Table 13. Summary Analysis of Variance of Mean First Person Pronoun/Total Words Ratios Source of Variation ss df MS F E_Inf1uence (A) 0.0000002 1 0.0000002 0.02 Incentive (B) 0.0000015 1 0.0000015 0.16 A x B 0.0000992 1 0.0000992 9.85* Error 125 0.0000091 9:2 < o 05 Hypothesis £§ Belief-discrepant messages encoded by persons with low 2 Influence under conditions of low justification will contain significantly longer sentences than belief-discrepant messages encoded in the other three conditions. The means and standard deviations for the words/sentence measure for subjects in the four experimental conditions are presented in Table 19. An analysis of variance yielded no significant effects. This lack of significance is partially due to the rather high variances in each of the four conditions. The pattern of means is consistent with Hypothesis 5; that is, low 2 Influence subjects in the Low Incentive condition wrote essays containing sentences that were some- what longer than the sentences contained in the essays of the other three experimental groups. However, since these differences were not significant, it must be concluded that the results fail to support Hypothesis 5. #6 Table 1n. Means and Standard Deviations of Words/Sentence Ratios Across Four Experimental Conditions 2_Influence High Incentive Low Incentive High 'x' = 19.30 19.02 s = 5.20 n.30 3 = 29 39 Low if = 19.3u 20.38 s = 0.35 1+.88 3 = 28 33 Hypothesis £3 Belief-discrepant messages encoded by persons with low n_Inf1uence under conditions of low justification will contain significantly more cross-outs and insertions than belief-discrepant messages encoded in the other three conditions. Table 15 contains a summary of the mean cross-outs and insertions/ total words for subjects in the four experimental conditions. It can be seen that high 2 Influence subjects writing under conditions of High Incentive and low £_Inf1uence subjects writing in the Low Incentive group made more errors (cross-outs and insertions) than subjects writing in the other two conditions. Table 15. Means and Standard Deviations of Cross—Outs and Insertions/Total Words Ratios Across Pour Experimental Conditions 3 Influence High Incentive Low Incentive High 32' = 0.0103 0.0090 S = 0.0153 0.0124 n = 29 39 Low 3? = 0.0070 00110 S = 0.0084 0.0133 2_= 28 33 47 An analysis of variance of the mean error ratios (Table 16) revealed a significant interaction between 2_Influence and monetary incentive. It is interesting to compare the error ratio results with the findings on use of first person pronouns. If one is willing to accept error rate as an indicant of cognitive stress. then the data presented in Table 15 indicate that subjects in the High 2.1nfluence- High Incentive and the Low 2_Influence-Low Incentive groups experienced more stress than subjects in the other two conditions. Moreover, if we consider the use of first person pronouns an indicant of willingness of a person to identify himself with his belief- discrepant arguments, then Table 12 indicates that subjects in the Low B'Influence-High Incentive and the High 2 Influence-Low Incentive groups identified themselves with their belief-discrepant arguments more than did subjects in the other two conditions. Thus. the findings obtained for Hypotheses 4 and 6 suggest that persons who experienced relatively more cognitive stress were less apt to identify themselves with their belief-discrepant arguments, while persons who experienced less stress were more willing to identify themselves with their arguments. Table 16. Summary Analysis of Variance of Mean Error Ratios Source of Variation ss df MS F 2 Influence (A) 0.0000050 1 0.0000050 0.96 Incentive (B) 0.0000003 1 0.0000003 0.05 A x B 0.0000216 1 0.0000216 H.15* Error 125 0.0000052 9'32 <005 Chapter 4 Discussion In general, the three hypotheses concerning attitude change, or self-persuasion, are only partially supported by the findings of the present study. The results obtained for Hypothesis 1 are particularly discouraging. Neither level of 2 Influence nor magnitude or monetary incentive were found to be related to attitude change after commitment to engage in belief-discrepant communication behavior. There are several possible explanations for this failure to confirm Hypothesis 1. Although the perceived adequacy of payment manipulation check revealed that persons in the 50 cent group rated their payment signi- ficantly less adequate than did persons in the $2.50 group, the mean rating of the 50 cent group was above the theoretic midpoint of the adequacy scale. Thus, persons in the 50 cent condition saw their payment as somewhat adequate. Comments made by subjects in the class debriefing sessions provide further insight into the adequacy of pay- ment problem. Several persons who received the 50 cent payment said that this sum was about as much money as they expected to receive; however, others remarked that they expected to receive more than 50 cents for their participation. Still other persons reported that 50 cents was not a bad wage for 10 minutes of work. These persons said that if the essay writing task had been longer, they would have con- sidered the 50 cent payment inadequate. A final possibility involves the spread of information about the experiment. Since all persons who received 50 cents for their essay writing were subsequently paid 48 09 an additional $2.00, it is possible that some persons assigned to the 50 cent condition learned of the additional $2.00 payment before they participated in the experiment. If a subject in the 50 cent group obtained such information, he might rate the 50 cent payment as quite adequate because of his expectations regarding the $2.00 payment. Other comments made in the debriefing sessions indicated that some persons justified their belief-discrepant communication behavior in still other ways. Several persons indicated that because the sponsor of the project was important, they agreed to do the essay writing task. In fact, a few individuals wrote comments on their final questionnaires stating that theywould have participated in the project even if a monetary payment had not been offered. Since 7” percent of the persons who agreed to participate in the project actually arrived even though they did not know how much they would be paid and they did not receive extra grade credit or other such incentives for their participation -- it appears that many persons thought the project was important. Some of the individuals in the 50 cent group said that although they were given the option of leaving the experiment after learning about the payment for the task, they stayed because they had expended considerable effort walking to the building where the experiment was conducted. Moreover, othenspointed out that they had reserved a certain period of time for participating in the project; thus, they felt that receiving 50 cents for the time spent was better than receiving no money at all. 50 It is evident from the pretest data reported in Chapter 3 that most students were extremely opposed to the idea of eliminating college student draft deferments. Several persuasion studies suggest that per- sons with extreme positions on an issue are less likely to change their attitudes than are persons with moderate positions. Thus, one would expect relatively little change in attitudes toward the elimination of draft deferments given the students' extreme initial positions. However, it should be noted that while there was little attitude change in the test of Hypothesis 1, significant changes in attitudes were observed in the tests of Hypotheses 2 and 3. These results would suggest that the manipulations used to test Hypotheses 2 and 3 were somewhat more effective than the manipulations used to test Hypothesis 1. Another possible explanation for the failure to confirm Hypothesis 1 concerns subjects' levels of ego involvement in the draft deferment issue. Prior self-persuasion experiments have employed such issues as, preventing the Ohio State University football team from going to the Rose Bowl, and banning known Communists from speaking at state— sponsored institutions. In contrast to these issueS, the draft defer- ment issue would seem to be more ego involving, since most college males 2253 face the possibility of being drafted. Females might also be highly involved in the draft deferment issue because the draft could interfere with their social life and their marriage plans. Data gathered by Rebach (1968) from Michigan State University students after the present study was completed indicate that the draft deferment issue is extremely ego involving. 51 It should be pointed out that if the above ego involvement explanation has validity, then a possible weakness of dissonance theory has been exposed. The results obtained for Hypohtesis 1 suggest that dissonance theory cannot accurately predict the relation— ship between justification and selfepersuasion at the simple commit- ment level, i£_the attitudinal issue used in the experimental test is highly ego involving. Of course, the preceding assertion assumes that factors other than ego involvement with the issue did not operate to a significant degree and that ego involvement was one of the chief reasons for the lack of attitude change in the test of Hypothesis 1. The validity of these assumptions rests upon a test of the ego involve- ment hypothesis. One final explanation for the failure to confirm Hypothesis 1 is the relatively low reliability of the 2 Influence measure. It will be recalled that the analyses of the 2_Influence measure indicated considerable disagreement between coder pairs; thus producing low intercoder reliabilities. It is possible that the effects predicted in Hypothesis 1 were not powerful enough to overcome this lack of reliability. In contrast to the results obtained for Hypothesis 1, the findings for Hypotheses 2 and 3 are encouraging. Support was found for the notion that high E_Influence individuals are more responsive to successful influence of others than are persons with low 2_Influence. Specifically, after taking a belief-discrepant stand, high 2 Influence persons who successfully persuaded others to adopt that position mani- fested greater self-persuasion than successful low 2 Influence 52 persuaders. By contrast, there is little difference in the amount of self-persuasion between high and low n_Inf1uence persons who failed to persuade. The difference in attitude change between the high and low 2 Influence persons who succeeded was large enought to result in a significant n_Inf1uence main effect; thUS confirming Hypothesis 3. The findings for Hypothesis 2 support those of earlier studies (Scott, 1957; Scott, 1959; Bostrom £3: 31., 1961; and Wallace, 1966) which found a positive relationship between reinforcement and magnitude of self-persuasion following belief-discrepant behavior. Moreover, the present findings further suggest that when successful persuasion is employed as a reinforcing contingency in the self—persuasion situation, successful high 2_Inf1uence persons will manifest more self- persuasion than successful low 3 Influence persons. It should be emphasized, however, that the findings for Hypothesis 2 are tentative at best and that further study is needed regarding the relationships among 2 Influence, influence feedback, and self-persuasion. In presenting the rationale for Hypothesis 3, it was suggested that persons with high 5 Influence would exert more effort at per- suading than would persons with low n_Inf1uence. Greater effort should lead to more dissonance; thus, persons with high n_Inf1uence should change their attitudes more than individuals with low 2.1nflu- ence. This prediction was confirmed. It was expected, however, that differences in effort expended by the high and the low 2 Influence groups would be reflected by variations in their message outputs. Since there were no significant main effects differences between 2 Influence groups in the tests performed for Hypotheses H through 6, 53 it is difficult to attribute the differences in attitude change to difference in effort on the part of the two 2_Influence groups. It should be noted that such criteria as essay length and encoding time could not be used as indices of effort in the present study. Con- sideration of these particular criteria might have shed further light on the effort hypothesis. While none of the message hypotheses (Hypotheses u through 6) received full support, the results indicate that high 2 Influence persons who encoded belief-discrepant messages under low justifica- tion conditions and low 3 Influence persons who encoded under high justification conditions experienced less cognitive stress and were more willing to identify themselves with their belief-discrepant arguments than were persons in the other two conditions. Contrary to the predictions made in Hypotheses u and 6, persons with high 2 Influence who received high justification manifested relatively more cognitive stress and were less willing to identify themselves with their belief—discrepant arguments than were high 2_Inf1uence persons who received low justification and low n_Inf1uence individ- uals who receiVed high justification. Perhaps the high 2_Influence persons who were paid $2.50 to write their belief-discrepant essays perceived their payment as recompense for persuading their readers, while the low 2 Influence persons who received the $2.50 payment saw it as a reward for merely writing the essay itself. Knowing that it is difficult to change others' attitudes with one message, the High Incentive—High n Influ- ence persons were under more pressure to persuade members of their 5“ audience than were the highs who received only 50 cents. Thus, the high 2 Influence persons who received $2.50 might have been "overly motivated" to persuade and more afraid of failure than the highs who received only 50 cents. This interpretation is supported by the fact that high n_Inf1uence persons in the high justification condition manifested a high level of cognitive stress and were less willing to identify themselves with their belief-discrepant arugments. The preceding interpretation is also partially supported by the findings concerning adequacy of payment. Persons in the High 2 Influence—High Incentive condition rated their payment less adequate than did Low n_Influence-High Incentive persons. It is possible that the high n_Influence persons' lower ratings of payment adequacy were due to their perceptions of the relationship between the payment and the task. Specifically, high n_Inf1uence individuals may have regarded the $2.50 as a somewhat less than adequate sum when the task was to persuade others to adopt an unpopular position on an important issue. The same $2.50 payment may have been perceived as relatively more adequate by the low annfluence group because they saw the payment as a reward for merely writing their assays. The preceding explanation of the results obtained in the tests of Hypotheses H and 6 for the High 2_Influence—High Incentive condi- tion exposes one of the chief problems in the justification-self— persuasion controversy discussed earlier. This problem centers on the rather vague nature of the justification concept itself. As indi- cated above, it is clear that persons who participated in the present study justified their belief-discrepant communication behavior in a 55 multitude of different ways. While some of these loci of justification might possibly been better controlled, it is exceedingly difficult to eliminate all possible sources of justification other than the one being manipulated by the experimenter. Moreover, when money is employed as a source of justification, further problems are introduced. For example, in the Carlsmith at: if: (1966) study, subjects who wrote belief-discrepant essays and subjects who deceived a confederate were all offered the same range of mone- tary rewards for task performance. However, a five dollar payment for writing a belief—discrepant essay may well be seen as a less adequate payment than a five dollar payment for face-to-face deception, since face-to-face deception may be perceived as a less difficult task than essay writing. In turn, these discrepancies between per- ceived task difficulty and amount of money offered to perform the task might possibly affect the subsequent magnitude of self-persuasion. It is obvious from the above discussion that the relationship between justification and self-persuasion after counterattitudinal behavior is a complex one. While the present study employed a rather elaborate design to study this complex phenomenon, an alternative strategy lies in performing a series of smaller, related studies which aSSess attitude change at different stages of the self-persuasion process. This approach actually might be preferable to the one taken in the present study, since several repeated measures on the same subjects within the same experiment generate sensitization problems which are difficult to control. 56 For instance, at least three relatively simple studies could be performed to determine more clearly the relationships among 2_Influence, influence feedback, justification, and self-persuasion. The first study could investigate the interaction between 2 Influence and justification on self-persuasion after simple commitment to engage in belief-discrepant behavior. The paradigm for this experiment would be similar to that Used to test Hypothesis 1 of the present study. A second study could then employ the same independent variables as those used in the first, but measures of self-persuasion would be obtained immediately following performance of the belief-discrepant act, for example, after encoding a counterattitudinal essay. This experiment would, perhaps, provide a better test of the relationship between 2 Influence and persuasive effort than was provided in the present study. Finally, a third study could be concerned with the possible inter- action between 2 Influence and influence feedback on self-persuasion after engaging in belief-discrepant behavior. This study would serve to provide further data relevant to Hypothesis 2 of the present study. In this particular study, justification would not be used as an independent variable. It was stated in Chapter 1 that the use of 2 Influence in the present study was considered exploratory. Now that this particular exploration is completed, the utility of the variable must be assessed. It was noted that the coding system for scoring n_Influence is somewhat complex; as a result, high interjudge reliabilities were difficult to 57 obtain. But in spite of this reliability problem, 2 Influence did a fairly good job of predicting in Hypotheses 2 and 3 and yielded some interesting interactions with monetary incentives in Hypotheses u and 6. These results suggest that 3 Influence is a useful variable for studying the self-persuasion phenomenon, particularly if a more reliable coding scheme can be developed. LIST OF REFERENCES References Aronson, E. The psychology of insufficient justification: An analysis of some conflicting data. In S. Feldman (Ed.), Cognitive consistency: Motivational antecedents and behavioral conse uents. New York: Academic Press, 1966, Pp. 109-133. Atkinson, J. W. (Ed.) Motives in fantasyI action, and society. Prince- ton: D. Van Nostrand, 1958. Bettinghaus, E. P. and Preston, I. L. Dogmatism and performance of the communicator under cognitive stress. Journalism Quarterly, 1969, 01, 399—002. Bostrom, R., Vlandig, J., and Rosenbaum, M. Grades as reinforcing contingencies and attitude change. Journal 2£_Educational Psychology, 1961, 52, 112-115. Brehm, J. W. and Cohen, A. R. Explorations in cognitive dissonance. New York: Wiley, 1962. Carlsmith, J. M., Collins, B. E., and Helmreich, R. L. Studies in forced compliance: I. The effect of pressure for compliance on attitude change produced by face-to-face role playing and anonymous essay writing. Journal 2f Personality and Social Psychology, 1966, '4, 1-130 Cohen, A. R. An experiment on small rewards for discrepant compliance and attitude change. In J. W. Brehm and A. R. Cohen, Explorations in cognitive dissonance. New York: Wiley, 1962, Pp. 73-78. Combs, W. H. The effect of audience feedback on encoding behavior of speakers. Unpublished masters' thesis, Department of Communication, Michigan State University, East Lansing, 1968. Elms, A. and Janis, I. L. Counter-norm attitudes induced by consonant versus dissonant conditions of role playing. Journal 2£_Experimental Research in_Personality, 1965, 1, 50-60. Festinger, L. A theory of congitive dissonance. Stanford: Stanford University Press, 1957. Festinger, L. and Carlsmith, J. M. Cognitive consequences of forced compliance. Journal 2£_Abnormal and Social Psychology, 1959, 58, 203-210. Greenberg, B. S. Performance and message consequences of encoding behavior under cognitive stress. Unpublished doctoral dissertation, University of Wisconsin, Madison, 1960. 58 59 Hovland, C. I. and Janis, I. L. (Eds.) Personality and persuasibility. New Haven: Yale University Press, 1959. Hovland, C. 1., Janis, I. L., and Kelley, H. H. Communication and persuasion. New Haven: Yale University Press, 1953. Hunt, M. P. and Miller, G. R. Open-and-closed-mindedness, belief- discrepant communication behavior, and tolerance for cognitive inconsistency. Journal of Personality 222 Social Psychology, 1968, 8, 35-37. """""""' "'"" " Insko, C. A. Theories of attitude changg. New York: Appleton-Century- Crofts, 1967. Janis, I. L. and Gilmore, J. B. The influence of incentive conditions on the success of role playing in modifying attitudes. Journal 25. Personality and Social Psychology, 1965, 1, 17-27. Janis, I. L. and King, B. T. The influence of role-playing on opinion change. Journal 2f Abnormal and Social Psychology, 1959, 99, 211-218. King, B. T. and Janis, I. L. Comparison of the effectiveness of improvised versus nonimprovised role-playing in producing opinion change. Human Relations, 1956, 9, 177-186. Linder, D. E., Cooper, J. and Jones, E. E. Decision freedom as a determinant of the role of incentive magnitude in attitude change. Journal 2f Personality and Social Psychology, 1967, 6, 295-259. McClelland, D. C. The achieving society. Princeton: D. van Nostrand, 1961. McClelland, D. C., Atkinson, J. W., Clark, R. A., and Lowell, E. L. The achievement motive. New York: Appleton-Century-Crofts, 1953. Miller, G. R. Variations in the verbal behavior of a second speaker as a function of varying audience responses. Speech Monographs, 1960, 31, 109-115. Rebach, H. Unpublished study. Department of Communication, Michigan State University, East Lansing, 1968. Rosenberg, M. J. When dissonance fails: 0n eliminating evaluation apprehension from attitude measurement. Journal 2: Personality and Social Psychology, 1965, 1, 28-92. Rosenberg, M. J. Some limits of dissonance: Toward a differentiated view of counter-attitudinal performance. In S. Peldman (Ed.), Cogpitive consistengy: Motivational_antecedents and behavioral conseqpents. New York: Academic Press, 1966, Pp. 135-170. 60 Scott, W. Attitude change through reward of verbal behavior. Journal 2£_Abnormal and Social Psychology, 1957, 55, 72-75. Scott, W. Attitude change by response reinforcement: replication and extension. Sociometry, 1959, 22, 328-335. Siegel, S. Nonparametric statistics for the behavioral sciences. New York: McGraw-Hill, 1956. Stolz, W. S. and Tannenbaum, P. H. Effects of feedback on oral encoding behavior. Language and Speech, 1963, 6, 218-228. Uleman, J. S. A new TAT measure of need for power. Unpublished doctoral dissertation, Department of Social Relations, Harvard University, Cambridge, 1965. Uleman, J. S. N influence, a new power—related motive; development and validation of a TAT measure. Paper presented at the meeting of the Midwest Psychological Association, Chicago, May 5, 1966. Veroff, J. Development and validation of a projective measure of power motivation. Doctoral dissertation, University of Michigan, Ann Arbor, 1955. Veroff, J. Development and validation of a projective measure of power motivation. In J. W. Atkinson (Ed.), Motives in fantasy, action, and society. Princeton: D. van Nostrand, 1958. Wallace, J. Role reward and dissonance reduction. Journal 2: Personality and Social Psychology, 1966, 3, 305-312. Winer, B. J. Statistical principles in eerrimental desigg. New York: McGraw-Hill, 1962. Zimbardo, P. The effect of effort and improvisation on self- persuasion produced by role-playing. Journal 2: Experimental Social Psychology, 1965, 1, 103-120. APPENDIX A Name Student # TEST OF CREATIVE IMAGINATION INSTRUCTIONS: This is a test of your creative imagination. Pour pictures will be projected on the screen before you. You will then have 60 seconds to look at the picture and then four minutes to make up a story about it. Notice that there is one page for each picture. The same four questions are asked. They will guide your thinking and enable you to cover all the elements of a plot in the time allotted. Plan to spend about a minute on each question. I will keep time and tell you when it is about time to go on to the next question for each story. You will have a little time to finish your story before the next picture is shown. Obviously there are no right or wrong answers, so you may feel free to make up any kind of a story about the pictures that you choose. Try to make them vivid and dramatic, for this is a test of creative imagination. Do not merely describe the picture you see. Tell a story about it. Work as fast as you can in order to finish in time. Make the stories interesting. 61 PICTURE 1. What is happening? Who are the people? What has led up to this situation? That is, what has happened in the past? What is being thought? What is wanted? By whom? What will happen? What will be done? 62 PICTURE 2. What is happening? Who are the people? What has led up to this situation? That is, what has happened in the past? What is being thought? What is wanted? By whom? What will happen? What will be done? 63 PICTURE 3. What is happening? Who are the people? What has led up to this situation? That is, what has happened in the past? What is being thought? What is wanted? By whom? What will happen? What will be done? 61+ PICTURE a. What is happening? Who are the people? What has led up to this situation? That is, what has happened in the past? What is being thought? What is being done? By whom? What will happen? What will be done? 65 APPENDIX B Name Student # ——_———— Opinion Profile Instructions The items contained in this booklet are concerned with your opinions and feelings about various matters. Please read each statement carefully, and then indicate your feeling about the statement on the scale provided. Place an "X" in the space which best represents your view on the statement. Remember, there are no right or wrong answers. Your opinions and feelings are what matter. Please make sure to print your name and student number at the top of this page. You may now begin. 66 l. 2. 67 The United States should pull all of its military forces out of Viet Nam immediately. Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly Military force is a useful instrument of national policy in some cases. Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly Draft deferments for college students should be eliminated. Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly 10. 68 We should pay more attention to our country's problems than to the problems of other countries. Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly The Republican and Democratic parties, as they now operate, do not adequately represent the wishes of the public in terms of candidate choice. Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly There is a need for a strong third political party in the United States. Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly If Robert Kennedy runs against Richard Nixon in the November election, Kennedy will probably win. Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly ll. 12. 13. l”. 15. 69 It is probable that Eugene McCarthy will be nominated by the Democratic party as its presidential candidate. Agree Strongly ,Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly Nelson Rockefeller of New York would probably beat Robert Kennedy in a presidential election. Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly The voting age should be lowered to 18 years of age. Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly College students, like non-college students under 26 years of age, should be subject to the draft at any time. Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly Communist China should be admitted to the United Nations. Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly 16. l7. 18. 19. 20. 70 The United States should decrease its military commitments overseas o Agree Strongly ‘ Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly There will be more riots in cities this summer than there were last summer. Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly Persons between the ages of 18 and 21 years know enough about political issues to make wise choices among political candidates in elections. Agree Strongly Agree Moderately Agree Slightly Don‘t Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly The United States should increase its military commitment in Viet Nam. Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly The war in Viet Nam will probably end within the next year. Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly 21. 22. 23. 2“. 25. 71 Laws should be enacted to prevent newspaper strikes such as those which have occurred in Detroit. Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly When foreign nations request U.S. military aid, the U.S. should refuse to render such assistance. Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly College students should be exempt from the draft while they are attending school. Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly Increased federal aid to poor persons will do little to prevent riots in cities. Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly Federal aid to education should be greatly increased. Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly APPENDIX C x2; Table I N Influence Scoring Categories, Brief Definitions, and Probabilities That Each Was More Frequent on Aroused TAT2 by Chance Category I Im (Influence Imagery) PR (Prestige) OR NS NR ND CR C O T :1: S US Total Score: (Organization) (No Self— Deprecation) (No Reminis- cence) (No Dread) (Counter- Reaction) (Consultation) (Threat) (Separation) Scored if... there are two people in the story who interact, so that one's action produces a reaction by the other. I Im is scored, and the story involves an important person, as indicated by his position, wealth, fame or glamour. I Im is scored, and an organization or hierarchy is involved in the story. I Im is scored, and there is no criti- cism, humiliation, or mention of per- sonal weakness of a character by him- self or author. I Im is scored, and there is no idle thinking about past experience. I Im is scored, and there is no fear or criticism of the future or future plans. I Im is scored, and one persons does something to another, who reacts in some way that relates back to the first person. I Im is scored, and people are consult- ing about the future, planning, or seek- ing and/or giving advice. I Im is scored, and one person threatens another's vital interests, so that he acts to counter that threat. I Im is scored, and two pe0ple are acts ively separated, leave each other, or are kept apart. P value* * 0.117 0.001 0.003 0.009 0.016 0.004 0.00” 0.001 0.027 0.005 Aroused subjects' total scores on TAT 'were greater than non-aroused scores at the 0.00016 level (l-tai Mann-Whitney U Test, N = 42). fiexact p values from the Wilcoxon average rank test, computed accord- ing to Klotz (Klotz,in press). 72 APPENDIX D /’— «'Ir‘z: Telephone Appointment Script Hello , I'm from the production staff of WMSB television. WMSB-TV and WKAR radio, in cooperation with the Television-Radio Department at Michigan State, are producing a series of programs dealing with students' problems. We are gathering material for these shows from a representative sample of students. Your name was chosen randomly as a part of this sample. We would greatly appreciate your cooperation. We hope that you will decide to participate, since we are trying to get a representa- tive and unbiased sampling of students' views. We are asking students to give us about 30 minutes of their time to help us obtain material for these programs. You will be paid for your participation. At this time I cannot tell you the exact amount you will be paid because we are still in the process of getting funds for the project. The project will be conducted in C-215 Wells Hall. Well Hall is located just across the Red Cedar River from the Computer Center. In order to make it more convenient for you to come, we have arranged seven different times during which you can participate. 73 APPENDIX E Name Student NO. WKAR RADIO CAMPUS POLL WKAR radio is conducting a student opinion poll on a randomly selected group of students. Your name was chosen as a part of this sample. It is important that you participate, since the data gathered in this poll will be used as an aid to future program development. The questions contained in this booklet concern your feelings and views about various issues. Also, there are some questions about your radio listening habits. Please answer all questions. Also, please put your name and student number in the spaces provided above. The data gathered in this poll will be reported on a group basis. Your answers will be kept in strict confidence You may now turn the page and begin. 74 75 First, we would like to ask you a few questions about yourself. 1. What was your age on your last birthday? 2. What is your current class standing? (Place an "X" in the appropriate box) D Freshman C7 Sophomore 5 Junior 3 Senior [3 Graduate Student 3. Are you: Male EU Female u. Do you intend to go on to graduate school? Yes DU No 5. How many radios do you have in your apartment or dormitory room? S None B One 5 Two 5 Three 3 a... or more 6. Do you own a portable radio? Yes DU No 10. 11. 12. About how much DUDUUU 76 time did you spend listening to the radio yesterday? Didn't listen at all Less than 1 hour Between 1 and 2 hours Between 2 and 3 hours Between 3 and u hours Four hours or more About how often would you say you listen to WKAR radio? DUDE About how many yesterday? HUNDRED When you listen to the radio, which one of the following stations Very often Sometimes Not very often Never news broadcasts did you listen to on the radio None One Two Three Four Five Six or more do you listen to most frequently? BURDEN Given that you WKAR WJIM WILS WVIC WJR OTHER Specify have free time, which one of the following activities do you prefer? DUDE Radio listening Television viewing Book reading Discussing issues with friends When you discuss problems with friends, do you try to convince them of your ideas, or do you mostly listen to their point of view? QUE Try to convince Mostly listen Do both equally 77 How do you feel about the following statements? 13. The sale of liquor should be legalized in East Lansing. Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly DDDDDDD la. All students at M.S.U. should be required to take at least a year of foreign language. Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly DDDDDDD 15. M.S.U. ought to raise its entrance requirements. Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly DDDDDDD 16. College students should be exempt from the draft while they are attending school. Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly DDDDDDD 17. Standards for admission to graduate schools are too high and should be lowered. Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly DDDDDDD 78 18. Student government at M.S.U. is relatively ineffective in terms of getting things done. Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly DDDDDDD 19. All students at M.S.U. should be required to take at least one year of mathematics and physics courses. Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly DDDDDDD 20. The United States should cease all bombing of North Viet Nam. Agree Strongly Agree Moderately Agree Slightly Don' t Know , Neutral Disagree Slightly Disagree Moderately Disagree Strongly DDDDDDD 21. College students, like non—college students under 26 years of age, should be subject to the draft at any time. Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly DDDDDDD 22. Which one of the following candidates do you think has the best chance of becoming the next President of the United States? George Wallace Robert Kennedy Richard Nixon Eugene McCarthy Nelson Rockefeller Hubert Humphrey , D DDDDD 23. 2Q. 25. 26. 27. 79 Which one of the following issues do you think is the most important in this year's election? Riots in big cities War in Viet Nam Control of atomic weapons Rising prices DDDD How justified were the actions of the students who rioted at Columbia University a few weeks ago? Very justified Somewhat justified Not at all justified DDD If the United States were to invade North Viet Nam, the Chinese Communits would probably enter the war. Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly DDDDDDD East Lansing merchants take advantage of students by charging them excessively high prices. Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly DDDDDDD Rents in the East Lansing area are excessive. Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly DDDDDDD 80 Students at M.S.U. are generally apathetic when it comes to taking action against such problems as high prices and high rents. Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly DDDDDDD Olin Health Center should dispense birth control pills to all female students who request them. Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly DDDDDDD Draft deferments for college students should be eliminated. Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly DDDDDDD In the future, the United States should not become involved in small wars like Viet Nam and Korea. Agree Strongly Agree Moderately Agree Slightly Don't Know, Neutral Disagree Slightly Disagree Moderately Disagree Strongly DDD DD APPENDIX F “\D Reader Reaction Sheet This tells how well you persuaded the students who read your script. The rating scales below tell you how the reader felt before he read your script and how he felt after he read your script. If, for example, the reader felt that the elimination of draft deferments for college students was a bad thing before he read your script, his check-mark appears on the bad side of the scale. If your script per— suaded him, his check-mark on the after reading rating should have mOVed to the good side of the scale; that is, after reading your script the reader changed his mind and now thinks that eliminating draft deferments would be a good thing. If you did not persuade the reader, his marks will be at the same position on both the before reading and the after reading scales. Elimination of College Student Draft Deferments Reader #1 Before Reading: Good : : ° ° ~ - Bad After Reading: Good : : ° ~ - - Bad Reader #2 Before Reading: Good : : : : : : Bad After Reading: Good : : ‘ ° ° ‘ Bad Reader #3 Before Reading: Good : : ' ° : ° Bad After Reading: Good : : : ' ' ° Bad Reader #u Before Reading: Good : ' ° : - - Bad After Reading: Good : ‘ ‘ ' ° ~ Bad 81 G x I D N E P P A WMSB—TV STUDENT OPINION SURVEY In an attempt to find out how students feel about several different problems, WMSB-TV is conducting this survey of student Opinion. Also, we are interested in students'.television viewing patterns, so that we can do a better job of programming. ' Please answer all of the questions in this booklet. Your responses will be reported on a group basis. Please put your name and student number in the space provided below. After recording your name and student number, please turn the page and begin. NAME STUDENT NUMBER _1. 82 83 1. Do you have a television set in your apartment or dormitory room? Yes No 2. About how much time did you devote to television viewing yesterday? None Less than 1/2 hour Between 1/2 hour and 1 hour Between 1 and 2 hours Between 2 and 3 hours More than 3 hours 3. How often do you listen to WILX/WMSB-TV (Channel 10)? Very often Sometimes Not very often Never 4. Between what hours do you watch television most on the average day? to 5. Which one of the following television stations do you think is most popular among students? WJIM (Channel 6) WILX/WMSB (Channel 10) WJRT (Channel 12) Other (specify; NOW, WHAT DO YOU THINK ABOUT THE FOLLOWING STATEMENTS? 6. Olin Health Center should dispense birth control pills to all female students who request them. Agree Agree Agree Don't Know Disagree Disagree Disagree Strongly Moderately Slightly Neutral Slightly Moderately Strongly 7. The United States should pull out of Viet Nam immediately. Agree Agree Agree Don't Know Disagree Disagree Disagree Strongly Moderately Slightly Neutral Slightly Moderately Strongly 84 8. The United StateSShouldnot be responSible for providing military aid to other nations. Agree Agree Agree Don't Know Disagree Disagree Disagree Strongly Moderately Slightly Neutral Slightly Moderately Strongly 9. Students should resist the draft by all means available. I n e I I o I Agree Agree Agree Don't Know Disagree Disagree Disagree Strongly Moderately Slightly Neutral Slightly Moderately Strongly 10. College students, like non-college students under 26 years of age, should be subject to the draft at any time. o o o o o n n Agree Agree Agree Don't Know Disagree Disagree Disagree Strongly Moderately Slightly Neutral Slightly Moderately Strongly ll. Open housing laws violate an individual's rights. / e I ‘ I I \ Agree Agree Agree Don't Know Disagree Disagree Disagree Strongly Moderately Slightly Neutral Slightly Moderately Strongly 12. Television coverage of riots helps to spread rioting throughout the country. Agree Agree Agree Don't Know Disagree Disagree Disagree Strongly Moderately Slightly Neutral Slightly Moderately Strongly 13. Early predictions of election winners on television influences the outcomes of elections. Agree Agree Agree Don't Know Disagree Disagree Disagree Strongly Moderately Slightly Neutral Slightly Moderately Strongly 1%. Violent programs on television cause some children to become criminals in later life. Agree Agree Agree DonTt’Know Disagree Disagree Disagree Strongly Moderately Slightly Neutral Slightly Moderately Strongly 85 15. The United States should invade North Viet Nam in order to shorten the war. Agree Agree Agree Don't Know Disagree Disagree Disagree Strongly Moderately Slightly Neutral Slightly Moderately Strongly 16. Students should have a greater voice in the making of university policy. Agree Agree Agree Don't Know Disagree Disagree Disagree Strongly Moderately Slightly Neutral Slightly Moderately Strongly 17. College students should be exempt from the draft while they are attending school. t o o 0 Agree Agree Agree Don't Know Digagree Disagree Disagree Strongly Moderately Slightly Neutral Slightly Moderately Strongly 18. The war in Viet Nam will end within the next six months. a o u o e o t Agree Agree Agree Don't Know DiSagree Disagree Disagree Strongly Moderately Slightly Neutral Slightly Moderately Strongly 19. Persons caught with marijuana in their possession should receive stiff jail sentences. Agree Agree Agree Don't Know Disagree Disagree Disagree Strongly Moderately Slightly Neutral Slightly Moderately Strongly 20. Michigan State University should totally eliminate football in order to improve its academic image. Agree Agree Agree DonTt Know Disagree DiSagree Disagree Strongly Moderately Slightly Neutral Slightly Moderately Strongly 21. LaWS should be enacted to ban anti-war demonstrations. Agree Agree Agree Don‘t Know Disagree Disagree Disagree Strongly Moderately Slightly Neutral Slightly Moderately Strongly 22. No restrictions should be placed on students visiting members of the opposite sex in dormitory rooms at any time. Agree Agree Agree Don't Know Disagree Disagree Disagree Strongly Moderately Slightly Neutral Slightly Moderately Strongly 86 23. Physical force should be used to prevent riots. o a o 0 u o ' Agree Agree Agree Don't Know DiSagree Disagree Disagree Strongly Moderately Slightly Neutral Slightly Moderately Strongly 24. Draft deferments for college students should be eliminated. a c o o Agree Agree Agree Don't Know Disagree Disagree Disagree Strongly Moderately Slightly Neutral Slightly Moderately Strongly 25. Persons who avoid the draft should receive long jail sentences. 0 a e t o I Agree Agree Agree Don't Know Disagree Disagree Disagree Strongly Moderately Slightly Neutral Slightly Moderately Strongly Finally, we would like to have your evaluation of this method of gathering material for programs. 26. How adequate was the money payment for the script writing part of the project? Very adequate Somewhat adequate Somewhat inadequate Very inadequate 27. Based on the readers' reactions to your script, how persuasive would you say your script was? Very persuasive Somewhat persuasive Not very persuasive Not at all persuasive 87 28. How much do you think your written script influenced the readers' opinions? __ Influenced them a lot ____._ Influenced them somewhat ______ Influenced them a little _____Didn't influence them at all 29. How enjoyable was it to take part in this project? Very enjoyable Somewhat enjoyable Somewhat unenjoyable Very unenjoyable 30. What do you think is the purpose of this project? THANK YOU FOR YOUR COOPERATION “NDINN