ATTITUDE CHANGE AND SELF-ATTRIBUTION OF RESPONSIBILITY AS FUNCTIONS OF ATTRIBUTIONS OF OTHERS BY WILLIAM D. STINNETT A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1977 ACKNOWLEDGMENTS Among those who contributed to the completion of this task are those who treated me well, less well, and not well at all. Those in the latter two groups I thank for making me tougher than I was . Those in the former group deserve more thanks than I could ever express. I can name but a few of the many who treated me well, although I love them all. Thanks to . . . The members of my committee: Judee Burgoon for her continuous support, her meticulous criticism, and her devotion to excellence; Thomas Saine for his thoughtful advice and counsel, especially in those difficult early stages of preparation; Barry Schlenker for his insightful questions which were always embarrassingly difficult for me; Marvin Shaw for being a model of scholarship and integrity . The chairman of my committee: Michael Burgoon, who has taught me much--love, hate, respect for my profession, and respect for knowledge--most of all for being my teacher, in a profession for which I have the highest regard. My other friends: Pamela Monast, who tolerated much but cared nonetheless, and who could always make me laugh at my own foolishness; Chuck Montgomery, who has always been my friend despite himself and others (I hope he finds what he wants); Michael Miller, rowdy Arkansas boy, who spoke seldom but wisely and who is both a realist and an idealist; Marshall Cohen, who kept his head and helped me keep mine through it all; Doug Vaughn, who was an amiable office companion, a true friend, and who provided an abundance of intellectual stimulation; Norm Markel, who always had a "dumb" joke to tell when I needed it most; Marc Reiss, whom I never understood but liked nonetheless; all of those poor souls who spent hours helping me prepare experimental materials and administer the experiment; the members of the VRT, who are rogues, comrades, and scholars. My friends and colleagues at Arizona State University: Bill Arnold, my boss, who said to me, "Finish your disser- tation this summer or you'll be out on the street"; the others in Arizona who pestered me an entire year to finish; my graduate students, who are first-rate friends, teachers, and researchers, and who made me look good by demonstrating their enthusiasm and competence in so many ways. My special thanks go to my parents to whom this work is dedicated. Their love and support was given without question or obligation. They gave so much and asked so little! TABLE OF CONTENTS ACKNOWLEDGMENTS LIST OF TABLES ABSTRACT CHAPTER I. STATEMENT OF THE PROBLEM Introduction Review of Research Rationale CHAPTER II. EXPERIMENTAL PROCEDURES Overview Subjects and Materials Procedure CHAPTER III. RESULTS Manipulation Check . Attitude Change Self-Attribution of Responsibility CHAPTER IV. DISCUSSION APPENDIX I. PRETEST QUESTIONNAIRE APPENDIX II. EXPERIMENTAL MATERIALS APPENDIX III. POSTTEST QUESTIONNAIRE REFERENCES BIOGRAPHICAL SKETCH . . Page ii v vi 1 1 4 20 25 25 25 32 34 34 3 5 41 43 53 6 6 72 7 4 80 LIST OF TABLES Table Page 1. Means and analysis of variance of scores for choice manipulation 35 2. Means and analysis of variance of experimental groups for semantic differential change scores 36 3. t tests of experimental comparisons for semantic differential change scores 37 4. Means and analysis of variance of experimental groups for Known Interval Scale change scores . 38 5. t tests of experimental comparisons for Known Interval Scale change scores 38 6. Dunnett's t test for control mean with experimental means for semantic differential change scores 39 7. Dunnett's t test for control mean with experimental means for Known Interval Scale change scores 40 8. t tests between control means for semantic differential and Known Interval Scale 41 9. Means and analysis of variance for self- attribution of responsibility scores 42 10. t tests of experimental comparisons for self- attribution of responsibility scales 42 Abstract of Dissertation Presented to the Graduate Council of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy SELF-ATTRIBUTION OF RESPONSIBILITY AND ATTITUDE CHANGE AS FUNCTIONS OF THE ATTRIBUTIONS OF OTHERS By William D. Stinnett August, 1977 Chairman: Michael Burgoon Major Department: Speech Personal responsibility for consequences has been offered as an explanation for the attitude change of subjects who perform counterattitudinal behavior. In forced-compliance experiments subjects are induced to engage in behavior which is contrary to their existing attitudes. If the circumstances are such that the subject cannot justify the behavior, he will change his attitude to correspond with the behavior in question. According to Cognitive Dissonance Theory, an uneasy or unpleasant feeling is associated with the inconsistency of having contradicted one's own belief. It is the desire to reduce or eliminate cognitive dissonance which motivates the person to seek some means of justifying the action. If no adequate justification can be found, the only avenue of dissonance reduction available to the subject is attitude change. Research has demonstrated that perceived choice, high effort, and public commitment contribute to dissonance-produced attitude change following counterattitudinal advocacy. Recent researchers have posited that these variables contribute to attitude change by increasing the subject's self- attribution of responsibility. In the present experiment it was reasoned that if subjects who wrote counterattitudinal essays believed they had influenced a person to change an attitude and that that person held the subject responsible for that attitude change, the subject's dissonance and feeling of personal responsibility would be increased and hence demonstrate greater attitude change. There were five stages to the experimentation in this study. First, subjects' initial attitudes were measured. Then the subjects were induced to write messages which were counter to their pretest attitudes. At a later experi- mental session, subjects received bogus feedback about the consequences of their messages. They were told by the target of the message that their messages were instrumental in changing the attitude of the target. Simultaneously, they received information from the recipient of the message about the degree of respon- sibility attributed to the subject by the target. Immediately following this information, the subjects were asked to respond to attitude items and items measuring felt responsibility. The results of this study were analyzed through a series of analyses of variance which demonstrated that the attribution of responsibility when communicated to the actor by the recipient of the persuasive message does have an impact on the actor's feeling or responsibility for the outcome of the action and the attitude of the actor toward the issue. Attributional messages from a target which hold that the actor behaved purposefully are given greater credence by the actor than other attributions of responsibility. In this study, subjects who received messages which contained purposive commission and justi- fied commission attributions reported a greater magnitude of felt responsibility and greater attitude change than all other subjects. CHAPTER I STATEMENT OF THE PROBLEM Introduction Investigation of the factors which contribute to attitude change following counterattitudinal advocacy has prompted considerable research. The amount of justifica- tion for an act, the aversiveness of the consequences of the act, the degree of commitment, and the amount of choice involved in performing the act have received much attention in the literature (Festinger, 1957; Brehm and Cohen, 1962; Aronson, 1968; Nel, Ilelmreich, and Aronson, 1969). More recently, researchers have posited that the responsibility felt by the actor of counterattitudinal behavior is the primary predictor of attitude change in forced-compliance situations (Brehm and Jones, 1970; Cooper, 197.1; Collins and Hoyt, 1972; Cooper and Goethals, 1974; Reiss and Schlenker, 1977). Forced-compliance experi- ments involve inducing subjects to engage in behavior which is contrary to their existing attitudes. If the circum- stances are such that the subject cannot justify the behavior, he will change his attitude to correspond with the behavior in question. According to Festinger (1957), there is an uneasy or unpleasant feeling associated with the inconsistency of having contradicted one's own belief. Festinger labels this feeling cognitive dissonance. It is the reduction or elimination of dissonance which moti- vates the person to seek some means of justifying the action. However, if no adequate justification can be found, the person will change the previously held attitude to correspond with the behavior. Attribution theorists con- tend that people are naturally motivated to observe and explain behavior. In one sense they function as amateur scientists and attempt to explain actions of individuals. The inferences people make about the intention, motives, responsibili ty, and abilities of individuals performing actions are the substance of attribution experiments (Heider, 1958; Kelley, 1971; Jones and Davis, 1965). In addition to explaining behavior in general, the principles of attribution theory apply to the explanation of one's own behavior. Miller and Ross (1975) suggest that experi- mental differences reported in the attributions of actors and observers may be explained by the actor's access to additional information about his own motivation and past behavior rather than any fundamental differences in the attribution processes. Bern (1967) contends that a person is an observer of his own behavior and judges his own actions in a manner similar to the way he evaluates others. Also, these attributions about oneself may be in- fluenced by the transmission of social cues. Festinger's Social Comparison Theory (1954) suggests that evaluation of one's own behavior is accomplished by comparing one's own acts with the behavior of significant others. Informa- tion obtained by observing the behavior of others may be used in making self-attributions. Rosenthal (1963) demon- strated that the evaluation of one's own behavior is influenced by the expectancies of others, presumably com- municated to the individual in very subtle ways. Schacter and Singer (1962) concluded that people may explain their own feelings by interpreting the actions of those around them. Given that the judgment of one's own behavior is influenced by the attributions of others, it is reasoned that persons who experience dissonance following counter- attitudinal advocacy will find it difficult to reduce that dissonance if they are held maximally responsible for their actions by those affected by those actions. There is a great deal of research that clarifies specific relationships between cognitive dissonance and attribution of responsibility. Review of Research Cognitive Dissonance Festinger (1957) maintains that any two cognitions may be consonant, dissonant, or irrelevant to one another. If the cognitive elements are in a dissonant relationship, as in the counterattitudinal advocacy situation (I am an honest, decent person. I am advocating a position which is not my own and may cause harm. ) , the individual is motiva- ted to reduce or eliminate the dissonance. If the behavior is irreversible and dissonance cannot be eliminated, it may be reduced by employing various strategies of rationaliza- tion (e.g. the subject may feel that he had no choice about performing the counterattitudinal act). The indi- vidual may feel that the act was justified as a result of extreme threat or a worthwhile reward. However, if no avenues of justification are open, the pressure will be to change his attitude to correspond with the behavior. Factors which are presumed to influence the existence and magnitude of dissonance are choice, commitment, effort, outcome valence and magnitude, competence and justification. Common research strategies for studying the relationship of these variables to dissonance arousal and subsequent reduction often employ a forced-compliance paradigm. in this model, subjects are provided minimal justification (reward or threat) to engage in behavior which is contrary to a presently held belief or attitude. If the behavior is irreversible and undeniable and the consequences of compliance severe, the subject will experience a psychologi- cal discomfort called cognitive dissonance (Festinger, 1957). Of the factors which have been demonstrated to increase dissonance, aversive outcomes and choice are most relevant to the present discussion. Choice. Dissonance, a post-decisional phenomenon, can be aroused only when the subject feels he had a choice in regard to performing the counterattitudinal behavior. Given that a person was given a choice, he will demonstrate more attitude change when given only a small reward than when given a large one. Presumably, the large reward serves as a means of justifying the counterattitudinal behavior, hence reducing the dissonance and eliminating the need to change his attitude (Festinger, 1957). This inverse relationship between attitude change and amount of incentive was obtained only for subjects who were given an opportunity not to comply with the experimenter's request in an experiment by Brehm and Cohen (1962). Similar dissonance results were reported for high-choice subjects in subsequent research (binder, Cooper, and Jones, 1967; Conolley, Wilhelmy, and Gerard, 1968; Holmes and Strickland, 1970; Sherman, 1970; Bodaken and Miller, 1971; Goethals and Cooper, 1972; Calder, Ross, and Insko, 1973; Pallak, Sogin, and Van Zante , 1974). Dissonance effects were produced by unattractive sources only for high- choice subjects in several studies (Smith, 1961; Powell, 1965; Zimbardo, Weisenberg, Firestone, and Levey, 1965). Subjects who chose to listen to a counter-persuasive message changed their attitudes in the direction of the message more than those who were not given a choice in a study by Jones et al. (1968). Eagly and Whitehead (1972) demonstrated that messages which were threatening to the subject's self-concept were more persuasive when sub- jects had chosen to listen. The subject's perception that he chose to engage in the counterattitudinal behavior has received considerable empirical support as a necessary condition for attitude change in forced- compliance situations. Aversive consequences. While choice lias been shown to be necessary for the production of dissonance, other research and interpretations indicate that it is not sufficient. Aronson (1968), Nel , Helmreich, and Aronson (1969), and Cooper and Worchel (1970) support the notion that a subject given high choice in counterattitudinal situations will experience cognitive dissonance and re- lated attitude change only if he believes the consequences of his behavior will be negative. Thie reasoning is sup- ported and expanded by Carlsmith and Freedman ( 1968) , Cooper (1971), and Goethals and Cooper (1975) who demonstrated that foreseeability of negative consequences is also necessary to produce attitude change in the counterattitudinal advocacy paradigm. According to Cooper's analysis, sub- jects can justify their behavior and hence reduce the dis- sonance by maintaining that they could not have known that their actions could have produced the negative consequences Brehm and Jones (1970) found that if subjects were sur- prised (consequences unforeseen) by the consequences of their behavior there was not corresponding attitude change. Responsibility for consequences. The results of two studies in which unforeseen negative consequences did pro- duce attitude change (Brehm, 1959; Sherman, 1970) are reinterpreted by Pallak, Sogin, and Van Zante (1974) in a series of studies in which subjects demonstrated atti- tude change following unforeseen negative consequences only if they could not attribute the outcome to chance. They say that "the task evaluation effect was replicated when the negative consequences were explicitly linked to the subject and his initial decision and was attenuated when the negative consequences were explicitly linked to chance and not to the subject" (p. 225) . Choice, self- attribution of responsibility, and valence of outcome were investigated by Arkin, Gleason, and Johnston (1976). They demonstrated that in the case of unexpected negative outcomes, subjects attributed responsibility to them- selves only if they felt they had a choice in their action. Several researchers have posited that responsibility rather than choice or negative outcome is the primary predictor of attitude change in counterattitudinal advo- cacy research. Brehm and Jones (1970, p. 4 31) state the following : It seems safe to conclude that Brehm and Cohen (1962) were quite wrong in hypothesizing that choice (volition) is a sufficient condi- tion for subsequent consequences to affect the magnitude of dissonance. The suggestion that a person must feel responsible for consequences appears to offer a better way of understanding the relevant experimental data. Since surprise was offered as a means for reducing dissonance, this was suggested as an explanation for subjects in the Brehm and Jones (1970) study not changing their attitudes. It was concluded that subjects assessed their own responsibility for consequences based on the foreseeability of the outcome. This analysis is con- sistent with that of Cooper (1971), who maintains that "a person will experience cognitive dissonance only to the extent that he feels responsible for his discrepant behavior and the consequences of that behavior" (p. 554). According to Cooper, this personal responsibility is due to the combination of choice and foreseeability. Goethals and Cooper, 1975) further demonstrated that even if the nega- tive consequences are subsequently eliminated, the dissonance is not then reduced unless the subjects were able to foresee that the aversive outcome would be eliminated. Pallak, Sogin, and Van Zante (1974) reasoned that perceived causality, in addition to choice and foresee- ability, is also a necessary condition for attitude change following counterattitudinal advocacy. They maintain, "Within the dissonance framework, variables other than foreseeability may determine responsibility for conseqeunces by determining responsibility for an initial decision" (Pallak et al., 1974, p. 217). They claim support for their rationale. Presumably negative consequences would result in task enhancement only when subjects made an internal attribution of causality for consequences and had high choice in the initial decision, in short, when subjects felt responsible for both the initial decision and for negative consequences resulting from the decision, (p. 224) While these authors interpret their data as indicating felt responsibility for both the initial decision and resulting negative consequences, they have established only that subjects made internal attributions of causal- ity for the decision and consequences rather than re- sponsibility. "Minimally, defining responsibility in terms of attributions of causality for consequences clarifies the conditions under which post-decisional consequences may result in positive attitude change" (Pallak et al., 1974, p. 226) . This is, however, an incomplete analysis of the role of responsibility in dissonance. Heider (1958, p. 112) claims in his discussion of personal 10 responsibility, "it is the intention of a person that brings order into the wide variety of possible action sequences by coordinating them to a final outcome." Hence, the concept of responsibility is pertinent only to the outcome of one's behavior not to the decision to perform the behavior. Many factors may influence one's decision, but, by definition, one can feel responsible only for the consequences of that decision. The Pallak et al. (1974) analysis is useful because it establishes the necessity for demonstrating a causal link between chosen behavior and negative consequences, but the meaning of responsibility is not clearly determined. Collins and Hoyt (1972) argue that high responsibility for aversive consequences is the most powerful predictor of the negative incentive results of forced compliance experiments. According to Collins and Hoyt (1972, p. 570) , it can be argued that in almost all studies reporting the dissonance-predicted, negative relationships the subject (1) assumed personal responsibility for his act and (2) felt that his act has serious consequences. Rather than manipulating choice, Collins and Hoyt (1972) told subjects that they were responsible or not responsible for the outcome of their counterattitudinal essays by stating the following: "You are, of course, responsible for the effects your essay may have (we want 11 to let you know that, although you are, of course, in no way responsible for the effects your essay may have) " (Collins and Hoyt, 1972, p. 573). In addition to the verbal instructions, subjects also signed a receipt, "I have chosen to write an open visitation essay for the U.C.L.A. Policy Evaluation Committee (Historical Records Committee) and hereby acknowledge receipt of 50C (2.50). Responsibility for its contents is mine. (I am in no way responsible for its contents)" (Collins and Hoyt, 1972). Also manipulated orthogonally were high or low conse- quences and high or low inducement (50C or $2.50) . The results clearly support the dissonance predictions for high responsibility, high consequences, and low inducement. While the Collins and Hoyt (1972) argument that the pesonal responsibility for consequences construct offers the best available integration of the forced compliance literature, they still have not offered a complete explana- tion of the determinants of responsibility in dissonance experiments. They have shown that one can produce dis- sonance predictions by substituting responsibility inductions in lieu of choice manipulations. Presumably in the Collins and Hoyt study had some subject not received a choice, he would have had an opportunity to deny his responsibility. Very likely, many subjects would also have been reluctant to sign the receipt in the high 12 responsibility condition if they felt they were being forced. Although they have demonstrated that among subjects given a choice, those willing to accept responsibility for their actions will demonstrate attitude change, the mechanism whereby responsibility is determined remains unclear . Reiss and Schlenker (1977) demonstrated that the inability to deny responsibility in forced-compliance- type situations is an important determinant of attitude change. Subjects given high initial choice for engaging in counterattitudinal advocacy with negative consequences demonstrated increasing amounts of attitude change as responsibility became increasingly more difficult to deny. In this study three observers (confederates) indicated to the subject whether or not they felt the subject has a choice about the decision to perform counterattitudinal behavior. The degree of observer agreement on the subject's decision freedom was presumed to determine the degree of difficulty in denying responsibility. While highlighting the importance of responsibility in forced-compliance situations, the Reiss and Schlenker (1977) study still leaves the question of the particular determinants of responsibility unanswered. 13 Responsibility While many of the studies cited have referred to the concept of responsibility as a predictor of attitude change in dissonance experiments, the determinants of personal responsibility have not been fully explored. Dissonance- related attitude change has been attributed to the sub- ject's felt responsibility in instances when the experi- mental manipulations were choice, foreseeabi lity , perceived causality, or some combination of these variables. A more thorough analysis of the components of responsibility would seem prudent before personal responsibility is offered as the major explanation for attitude change follow- ing dissonance. Levels of responsibility. Piaget (1932) has de- scribed responsibility developmentally . As a person matures, he considers an increasing number of factors in assessing personal credit or blame for the outcome of events. Heider (1958) views this as a five-stage process. He maintains that people assess the degree of personal responsibility for the outcome of incidents according to the levels of responsibility. The most basic factor considered in attributing responsibility is global association (Heider, 1958). Level one is defined as a situation in which a person is held responsible for the outcome of an event if he is 14 merely associated with the event. "At the most primitive level the concept is a global one according to which the person is held responsible for each effect that is in any way connected with him or that seems in any way to belong to him" (Heider, 1958, p. 113). In this situation, causality, f oreseeability , inten- tion, or justification is not considered in the assessment of responsibility. If one's neighbor, a Republican, were held responsible for the actions of the Republican party, it would be an example of global association. If, in addition to association, one is held respon- sible for consequences caused but not foreseen, intended, or justified, this is the second level which Heider labels extended commission. "Causation is understood in the sense that p was a necessary condition for the happening, even though he could not have foreseen the outcome however cautiously he had proceeded" (Heider, 1958, p. 113) Obviously, if the outcome had been foreseeable, the actor is viewed as negligent in the instance of negative out- comes. This parallels Heider 's third level of respon- sibility which tie calls careless commission. The most responsibility, of course, is assigned to those causing events who have had the opportunity to foresee the outcome and acted intentionally. Thus, according to Heider, the fourth level (purposive commission) carries the maximum amount of responsibility. The fifth level (justified commission) includes environmental restraints or pressures which share in the responsibility for the outcome. "We mean by this that anybody would have felt and acted as he did under the circumstances" (Heider, 1958, p. 114) . A study by Shaw and Sulzer (1964) offers empirical support for the viability of Ileider's framework for deter- mining responsibility. The researchers tested both adults and children by having them assess the responsibility of actors in descriptions of hypothetical situations repre- senting the levels of responsibility. The results were interpreted as supporting the notion of discriminating attributions or responsibility among adults and less discrimination among children. Global association and extended commission do not constitute sufficient circum- stances for the attribution of responsibility in adult populations. Responsibility is assigned to those who should have foreseen the consequences, and the most re- sponsibility is assigned to those who act intentionally. The expected decrease in responsibility attribution for justified commission did not receive unqualified support (Shaw and Sulzer, 1964). Both adult and juvenile popula- tions demonstrated less attribution of responsibility at level five for negative outcomes but not for positive. 16 Sources of felt responsibility. The impact of the communication of perceptions and judgments of others on one's behavior has received attention in several lines of research. Bern (1967) in his elaboration of self-perception theory contends that a person is an observer of his own behavior; therefore before one can feel responsible for an act, one must at least be able to observe oneself as the cause of that act. Festinger's social comparison theory (1954) suggests that evaluation of one's own behavior is accomplished by comparing one's own acts with the behavior of significant others. Rosenthal (1963) indicates that school children evaluate themselves and behave according to teacher expectancies presumably communicated to them by subtle social cues. It is also maintained that subjects in experiments respond to the situational demand character- istics inadvertantly (or deliberately) communicated to them by experimenters (Rosenthal, 1964). Several intriguing studies (Schacter and Singer, 1962; Nisbet and Schacter, 1966; Valins, 1966; Valins and Ray, 1967; Storms and Nisbett, 1970) support the notion that people use information from their social environment to identify the exact nature of internal states of arousal (such as anger, euphoria, sexual arousal, etc.). If cognitive dissonance may be viewed as a state of arousal, 17 it is reasonable that individuals will look for cues from their environment to explain the feeling. The process of attribution has received considerable attention by researchers (Heider, 1958; Kelley, 1971; Jones and Davis, 1965). Pittman (1975) supports the notion that dissonance may be realistically viewed as a state of arousal by demonstrating that if subjects who perform counterattitudinal advocacy are able to attribute their uneasiness to a source (viz., fear of shock) other than their inconsistent behavior, the dissonance will be medi- ated. Receiving less attention, however, is the impact of attributions on the behavior of those making the attribu- tions and the behavior of those to whom the attributions are communicated. Munson and Kiesler (1974, p. 453) main- tain that "a substantial area which has been relatively unexplored is the case where the attribution is made by one individual to another" (p. 453). This is the situation in which the inference of one individual is communicated to the actor. Munson and Kiesler (1974) argue that the communication of attributions would be more potent than either alone. Although they did not obtain the attribution- persuasion interaction which they predicted, a main effect for type of strategy supported the claim that the attri- bution strategy had a persuasive impact of its own. Miller, Brickman, and Bolen (1975) found that an attribution strategy was superior to persuasion in elicit- ing desired behaviors of second graders on two separate tasks. They explain the outcomes by suggesting "that persuasion often suffers because it involves a negative attribution (a person should be what he is not) , while attribution generally gains because it disguises persuasive intent" (p. 430). While the author does not accept the notion that the communication of an attribution, positive or negative, to elicit desired behavior is a fundamentally different process from persuasion, the implications of the strategy are of interest. In the Miller et al . (1975) study, the children who were told that they were neat and tidy or motivated to do well in math performed better on their respective tasks than those who were told that they should be neater or more motivated. It could also be reasoned that telling individuals that they should be responsible for certain actions might be less effective than telling them that they are responsible. Rosenbaum and Zimmerman (1959) posited an external commitment effect. "If an external source attributes to an individual a particular opinion prior to exposure to an attempt to effect change, effects similar to those produced by self-commitment will occur and can be de- scribed by the term 'external commitment'" (p. 247). 19 The effect was obtained only for congruent attributions (attributions in agreement with the subject's initial opinion) . The authors suggest that the pretest served as a prior commitment which might have precluded the effect of attitude change due to external, incongruent attribu- tions (attributions in disagreement with the subject's initial opinion) . The subjects had just unambiguously made their attitudes public, hence their susceptibility to incongruent suggestions should not be expected to be great. In the case of attribution of responsibility, the prior response (statement of attitude) is not relevant to the subsequent attribution of responsibility. It is likely that one source of a person's feeling of responsibility, in conjunction with the actual circum- stances of the event, is information received from other people. Such questions were posed by Reiss and Schlenker (1977, p. 3) . Naturally, audiences frequently provide an actor with their perceptions of whether or not the actor was personally responsible for the behavior. Given a discrepancy between a person's initial, perceptions of responsibility for an action that produces aversive conse- quences and the perceptions of observers, what happens? Does the individual follow his initial perceptions, does he follow those of the audience or does he compromise? This study demonstrated that information from observers does indeed influence subject's ratings which lends 20 support to the notion that attributions of responsibility communicated to the subject have an impact on his own feelings of responsibility. By extension, it would also follow that if the source of this information was the person most directly affected by the consequences of the subject's action, that the impact might be greater than for an observer (Cialdini, Braver, and Lewis, 1974; Touhey, 1971) . Rationale Many researchers agree that personal responsibility plays an important role in forced-compliance-related attitude change. Whether this responsibility operates on attitude change by increasing dissonance or whether the appearance of responsibility prompts face-saving strategies is less well understood. Impression management theorists suggest that reported attitude discrepancies in counter- attitudinal situations are due to the subject's desire to avoid the embarrassment of appearing inconsistent or insincere to those observing the behavior rather than an internal feeling of inconsistency. Such explanations may be ruled out in the present study, however, because the recipient of the counte cattitudinal message apparently sees only the message and would have no reason to believe that the subject was in any way insincere or inconsistent. There are no other observers who could conceivably form 21 impressions about the behavior of the subject. In either case it is probable that knowledge of the judgments of others about one's behavior is an important consideration. Reiss and Schlenker (1977, p. 3) interpret Goethals and Cooper (1975) by reasoning, "Thus, the actual consequences of the behavior are less crucial than is the appearance that a person is responsible for a potentially aversive action." One might also contend that the consequences of an action may include both the attitude change of the target and the judgments of the target. The recipient who judges the subject as responsible for the attitude change would be likely to view the subject as one who holds the advocated position. Indeed, such an attribution may be viewed as aversive to the subject. It is reasoned that, in addition to contributing to one's feeling of responsi- bility, such judgments about the subject's responsibility could also increase the aversiveness of the consequences, and therefore increase the magnitude of dissonance. Dis- covery that targets yielded or did not yield influenced persuaders' ratings of their targets (Cialdini , 1971). It was demonstrated that subjects who were successful in persuasive attempts evaluated the target more positively than subjects who were unsuccessful. Also yielders were rated as more intelligent than non-yielders by the per- suader (Cialdini et al., 1974). Responsibility 22 attributions communicated to the subject by persuaded targets should be especially aversive for the persuader when the persuasive message is counteratti tudinal . Hence, the greater the degree of responsibility attributed to the actor of a counteratti tudinal act by the recipient of the consequences of that act, the more dis- sonance the actor will experience and the more difficult it will be for the actor to avoid a feeling of personal responsibility for the outcome. Hence, subjects who perform counterattitudinal behavior under circumstances of maximum attributed responsibility should demonstrate maximal amounts of attitude change. According to Shaw and Sulzer (1964), adults will not hold a person responsible for an action if he/she is (1) only associated with the act (global association) or (2) was the cause of the act but could not have foreseen the consequences (extended commission). They also contend that adults will hold a person accountable if he is seen as the cause of that act and should have been able to foresee the consequences of such action (careless commission) . According to this analysis the maximum degree of respon- sibility for an act will occur if the person also attri- butes intention to the actor (purposive commission) . However, the actor's responsibility should be mitigated by environmental pressures, such as reward or punishment, 23 which partly justify the action (justified commission) (Shaw and Sulzer, 1964) . Hypotheses Thus, subjects who perform counteratti tudinal be- havior and find that the recipient of the consequences of their action attributes causality, foreseeability, and intention to them for the action should experience the most dissonance, find it the most difficult to avoid responsibility and therefore demonstrate the most attitude change. Subjects whose targets attribute careless commis- sion to them should experience less dissonance, less responsibility, and less attitude change than those with attributed purposive commission, but more than subjects whose targets have made only global association or extended commission attributions. In addition, subjects who receive attributions of justified commission from their targets should experience less dissonance, find it easier to avoid responsibility, and demonstrate less sub- sequent attitude change than subjects who receive pur- posive commission attributions. The preceding arguments lead to the following hypotheses: H, : Following counterat ti tudinoJ advocacy, subjects whose targets communicate purposive commission attributions to them will demonstrate greater attitude change than subjects who receive global association, extended commission, careless commission, or justified commission attributions. 2 4 H_ : Following coun teratti tudinal advocacy, subjects whose targets communicate purposive commission attributions Lo them will express greater felt responsibility than subjects who receive global association, extended commission, careless commission, or justified commission attributions. H • Following coun teratti tudinal advocacy, sub- jects whose targets communicate careless commission attributions to them will demon- strate more altitude change than those who receive global association or extended commission attributions from their targets but less than those receiving purposive commission attribu- tions . H : Following coun terattitudinal advocacy, subjects whose targets communicate careless commission attributions to them will express more felt responsibility than those who receive global association oj extended commission attributions but less than those receiving purposive commis- sion attributions. CHAPTER II EXPERIMENTAL PROCEDURE Overview The methods for testing the hypotheses included five steps. First, subjects' initial attitudes were measured. Then the subjects were induced to write messages which were counter to their pretest attitudes. At a later experimen- tal session, subjects received bogus feedback about the consequences of their messages. They were told by the target of the message th.it their messages were instrumental in changing the attitude of the target. Simultaneously, they received information from the recipient of the message about the degree of responsibility attributed to the subject by the target. Immediately following this informa- tion, the subjects were asked to respond to attitude items and items measuring felt responsibility. Subjects and Materials Sub j ects . The subjects for this experiment were selected from the introductory course in the Department of Speech at the University of Florida. There were 201 students from 17 different classes who completed the pretest 2r> 2h attitude questionnaire (see Appendix 1). One hundred and forty-five experimental subjects were selected from this population, based on their agreement with the key pretest attitude item. Pretest questionnad re. The pretest questionnaire was composed of 16 statements about campus and national topics. The pretest was labeled an attitude survey which was part of a research project concurrently in progress in the Speech Department at the University of Florida. Each statement was followed by four, seven-interval semantic differential- type scales and an 11-point known interval scale. The semantic differential scales are four which demonstrated high factor purity in the evaluative dimension according to Osgood, Suci, and Tannenbaum (1957). The Known Interval Scale (Burgoon, Burgoon, and Vaughn, 1977) consisted of 11 points with each point anchored and weighted by succes- sive interval scaling. Means and standard deviations were calculated for each item by use of both the semantic differential scales and the Known Interval Scale. A polarized mean was desirable in order to insure that most subjects would be creating counterattitudinal messages. A small standard deviation was desirable to demonstrate a high rate of initial polarization among subjects. The item which most closely corresponded to both of these criteria was "Dorms should have 24-hour visitation." Pretest results showed subjects 27 to be in high agreement with this statement with little variation. This procedure allowed selection of a single item for the experimental task which would allow the maxi- mum number of pretest subjects to argue counteratti tudinally . Experimental task. Based on the results of the pre- test analysis, 145 subjects who indicated agreement with the statement "Dorms should have 24-hour visitation" were asked to write short, persuasive essays supporting the restriction of dormitory visitation. Each subject received a letter (see Appendix 2) from a bogus organization, "Office of Communication and Public Relations," which asked them to participate in a program to determine the attitudes of potential University of Florida students on the issue of dormitory visitation. Subjects were told that their letters along with letters from other college students would be given to high school students who were planning to attend the University of Florida in the fall. Subjects were told that a sufficient number of letters supporting open visitation had already been obtained; thus they were being asked to write letters in opposition to 24-hour visitation in order that each potential student would receive letters on both sides of the issue. The subjects' memo stated, "Since we have already received the student para- graphs written in opposition to restricted visitation, we are asking that you write a paragraph in favor of this 2 8 policy. ... On the enclosed 'memo' please write the most persuasive paragraph which you can think of in favor of restricted visitation." In addition to the written instruc- tions, experimenters read the memo aloud to each class and requested that the subjects put their names on their letters. Although encouraged to write the letters, subjects who objected were assured that they did not have to comply. This was necessary in order to ensure that subjects felt they could choose not to write the essay. Thirteen sub- jects refused to participate. Another 40 subjects wrote in favor of open visitation. These subjects, of course, were not included in the final analysis. All essays were read by three expert raters to determine if they were actually written in opposition to 24-hour visitation. Only those essays in which all three raters agreed were included in the analysis. Experimental condit Ions . The five levels of respon- sibility (Heider, 1958) constituted the experimental con- ditions. Five bogus letters supposedly written by high school students attending a college orientation program were constructed to represent the levels of responsibility. Subjects in Condition I (Global Association) received a letter which stated, "Before I read the letters from the college students, I thought that 24-hour visitation was a good thing, but I've changed my mind. Now I don't think 29 it's such a good idea. L had the idea that it might be good before I saw the letters, but now I'm against it." Several considerations were important in the construction of the letter for Condition I: (1) it was necessary to state that there was an effect — the high school student had indeed changed an attitude concerning 24-hour visitation; (2) the change was caused by the letters from the college students, not specifically by the subject's letter; (3) the high school student attributed the responsibility for the change to the subject. According to Heider (1958) the attribution of responsibility by global association is the situation in which the person is held responsible for any effect that he is connecied with in any way. Therefore the idea that the high schooi student experienced attitude change as a result of all the letters from the college students but nonetheless holds the subject responsible, reflects the notion of global association. Condition II (Extended Commission) has to communicate to the subject that (1) .Attitude change was effected; (2) the subject's letter was the direct cause of the change; (3) the target realized that the subject could not have foreseen such consequences; and (4) the target holds the subject responsible for the attitude change. The bogus letter for Condition II stated, "I used to think that 24- hour visitation was good, but I've changed my mind. Even 30 though you couldn't know that your letter would make me change my mind and you didn't really mean to, it made me think that 24-hour visitation isn't such a good idea. Before, I thought it might not be a bad idea, but not now. And even though you couldn't have had any idea that your letter would make me think different about it, it sure did." The subjects in Condition II (Careless Commission) recieved a letter similar to that in Condition II, with the exception that the target stated that the subject should have been able to foresee that the letter would cause attitude change. This Letter stated, "Before I read your letter, I thought that 24-hour visitation was a good thing. Even though you didn't really mean to change my mind, when you wrote the letter you probably knew that it might. Now I think that 24-hour visitation isn't such a good idea. Before, I had the idea that it might be a good idea, but now I don't think so. Maybe you didn't mean to make me think like you even if you knew that would happen, but anyway now I'm against it too." Of special importance in this condition was the avoidance of any implication of intentionality on the part of the subject. In Condition IV (Purposive Commission) the concept of intentionality was introduced. In this condition, subjects read the statement, "Before I read your letter, I thought that 24-hour visitation was a good thing. I am 31 glad that you tried to get me to change my mind and that you realized your letter: would convince me. Now I think that 24-hour visitation isn't such a good idea. Before, I thought it might be good, but not now. I think you wanted me to think like you and knew that would happen, but that's OK. I'm glad you had to do that assignment because now I'm against it too." Each bogus letter also contained one of three sets of opening and closing remarks (see Appendix 2) designed to add realism to the statements and to reinforce the notion of responsibility attribution. The various opening and closing remarks were randomly assigned across all condi- tions . Each statement was examined by five expert raters who were all in agreement that the statements accurately represented the five levels of responsibility by Heider (1958) . Through consultation with several high school teachers the language used in the bogus letters v/as carefully worded according to the type of grammar and phrasing typically used by college-bound high school seniors. In addition, all letters were handwritten in different handwriting styles and various colors of ink in order to increase the subject's perception of authenticity. 32 Posttest questionnaire. Immediately following the bogus letters, subjects received the final questionnaire (see Appendix 3) which included the statement about 24-hour visitation followed by i he same four semantic differential- type scales and the Known Interval Scale. On a separate page were Likert-type scales to measure the subject's felt responsibility for changing the high school student's attitude and the subject.' s perceived choice in writing the counterattitudinal essay. Also included was an open-ended question designed to defect subject's suspicion of the experimental intent. Procedure Subjects who had indicated on the pretest administered during the first week of the quarter that they opposed 24-hour visitation were eliminated. Five weeks later, remaining subjects were asked to complete the experimental task. Their essays were then examined to determine com- pliance with the instructions. Those subjects who did not write in opposition to their previous attitude on dormi- tory visition were eliminated from further analysis. Remaining subjects were then randomly assigned to each of the five experimental conditions and one control condition. One week later the subjects received the bogus letter representing the appropriate experimental condition. 33 Control subjects received no letter but had previously written a counterattituclinal essay. Students who were absent on the day the counteratti tudinal essays were written served as a pre test-posttest only control group. Immediately following the distribution of the bogus letters, subjects in the five experimental conditions and two control conditions received the posttest questionnaire, All subjects were completely debriefed at this time and the purpose of the study was explained. CHAPTER III RESULTS One-way analyses of! variance were used to test for differences among treatments for attitude change and self- attribution of responsibility. Planned mean comparisons were done with simple t tests. Comparisons of control means with experimental means were assessed with the Dunnet's t test. A one-way analysis of variance was also used for the perceived choice manipulation check. A t test was used to compare the two control conditions: (1) subjects who wrote < ounterattitudinal messages but received no responsibility attributions (Cl) and (2) sub- jects who completed the pretest and posttest but received no experimental manipulations (C2) . The 'analyses of variance and t tests us>.'d in these analyses use formulas adjusted for unequal cell sizes. Man i pulation Check In order to demonstrate subject's perception of choice, a one-way analysis of variance was performed for the five experimental conditions and the primary control condition (Cl). In this analysis, the F value was not significant at the .05 ' evel (see Table 1). This indicates 3 4 3 5 that there were no significant differences in the subjects' perception of the amount: of choice in whether to write the counterattitudinal essays. This outcome indicates a successful choice manipulation given that differential perceptions of choice could have mitigated against confirma- tion of the predicted differences in attribution and attitude change. Table 1. Means and analyst: of variance of scores for choice manipulation Global Extended Careless Purposive Justified Control Association Commission Commission Commission Commission 2-62 2.27 2.78 2.60 3.36 2.54 N=13 N=15 N=l>: N=15 N=25 N=ll Source of Variance df Between 5 7.0 7 1.41 Within Total 5 7.0 7 1.41 91 144.14 1.58 96 151.21 Note: Choice was measured on a five-point Likert-type scale ranging from 1, strongly agree, to 5, strongly disagree. Attitude Change Experimental analys Ls of semantic differential . The analysis of variance for the semantic differential scores produced a significant r ratio (see Table 2). The t tests demonstrated significant differences between means for global association (GA) and purposive commission (PC), 36 Table 2. Means and analysis of variance of experimental groups for semantic differenl Lai change scores Global Extended Careless Purposive Justified Association Commission Commission Commission Commission 1.31 3.60 2.44 8.12 7.15 N=15 N=18 N=15 N=25 N=ll Source of Variance df SS MS Between 4 62 2.57 155.64 4.71 Within 78 2577.12 33.04 Total 82 3199.69 Note: Attitude change was measured with four seven-point semantic differential scales. These scales were summed for both pretest and posttest scores. The difference scores could range from 0 to 24. extended commission (EC) and PC, careless commission (CC) and PC, GA and justified commission (JC) , EC and JC, and CC and JC . All other comparisons were not significant (see Table 3). These i "suits indicate that subjects who received purposive commission attributions from their targets demonstrated more attitude change than all other experimental subjects except those receiving justified commission attributions. Also, those subjects receiving justified commission attributions demonstrated significantly more attitude change than all other experimental subjects except for those receiving purposive commission attribu- tions. The analysis of variance for semantic differential 37 scores accounted for 19 percent of the total variance . Table 3. t tests of experimental comparisons for semantic differential change scores 1 Comparison GA:EC 1.19 ns GA:CC .55 ns GA:PC 3.70 <.05 GA:JC 2.72 <.05 EC:CC .60 ns EC: PC 2.62 <.05 EC:JC 1.73 <.05 CC-.PC 3.09 <.05 CC:JC 2.20 <.05 PC:JC .49 ns GA=Global Association; EC=Extencled Commission; CC=Careless Commission; PC=Purposive Commission; JC=Justified Commission Experimental analysis of Known Interval Scale. The analysis of variance for the Known Interval Scale pro- duced a significant F ratio (see Table 4). Subsequent mean comparisons demonstrated significant differences between GA and PC, EC and PC, CC and PC (see Table 5). It should also be noted that with the exception of cell three, the pattern of means is as predicted. The analysis of variance for the Known Interval Scale accounted for 19 percent of the total variance. Table 4. Means and analysis of variance of experimental groups for Known Interval Sc;le change scores Global Extended Careless Purposive Justified Association Commission Commission Commission Commission 1-12 1.26 .85 3.06 2.32 N=14 N=18 N-16 N=24 N=ll Source of Variance df SS Between ; 87.56 21.89 3.90 <.05 Within Total 4 87.56 21.89 78 4 37.41 5.61 82 524.97 Note: Attitude change was measured by subtracting pretest from posttest scores on the 11-point Known Interval Scale. Weighted scale values ranged from .66 to 11.31. Table 5. t tests of experimental comparisons for Known Interval Scale change scores Comparison GA:EC .20 ns GA:CC .36 ns GA:PC 2.4 3 <.05 GA : JC 1.32 ns EC:CC .51 ns EC: PC 2.47 <.0 5 EC:JC 1.17 ns CC.-PC 4.71 <.05 CC:JC 1.59 ns PC:JC .86 ns Critical value for t (<.05, df=78) = 1.66. 39 Control comparisons. A one-way analysis of variance was performed which included the five experimental groups and the primary control group (Cl) for the semantic differ- ential scores. This was done in order to generate an estimate of pooled variance to use in the Dunnett's t test for comparisons with the control mean. Significant differences were found for comparisons of the control mean with PC and with JC. All other comparisons were not significant (see Table 6) . This result indicates that Table 6. Dunnett's t test for control mean with experimental means for semantic differential change scores Group Mean SD Dunnett ' s t Control 1.56 ■1.46 GA 1.13 4.78 .41 EC 3.60 5.71 1.58 CC 2.44 5.7 3 1.46 PC 8.12 5 . 66 10.4 7 JC 7.15 6.48 3.5 5 ns ns ns <.05 <.05 Critical value for t (<.05, df=6,86) = 2.58. subjects receiving high responsibility attributions (purposive commission and justified commission) from their targets demonstrated significantly more attitude change than subjects who received no attributions following their counterattitudinal essavs. There were no differences 4 0 among students receiving lower responsibility attributions (global association, extended commission, and careless commission) and those receiving no responsibility attribu- tions. Control comparisons for the Known Interval Scale produced a significant difference for control (Cl) vs. purposive commission. All other comparisons were not significant (see Table 7) . Table 7. Dunnett's t test for control mean with experimental means for Known Interval Scale change scores Group Control GA EC CC PC JC .27 1.12 1.26 .85 3 . 06 2.32 SD 1.80 3.11 2 . 26 1.50 2.79 2.82 Dunnett ' s t Critical value for t (<.05, df=6,86) = 2.58. .91 1.1] .51 3.29 2.06 ns ns ns <.05 ns A simple t test between the two control means (see Table 8) indicated no significant difference in attitude scores between those subjects who wrote counterattitudinal essays and received no feedback and those subjects who completed the pretest and the posttest but received no experimental manipulation. The t test between control means for the Known Interval Scale produced no significant differences (see Table 8). 41 Table 8. t tests between control means for semantic differential and Known Interval Scale Mean (CI) SD(C1) Mean (C2) SD (C2) Semantic 1.56 Differential N=ll Scales Known Interval .27 Scales N=ll 4.46 .48 3.98 .39 N=26 .18 1.45 .16 N = 26 Self-Attribution of Responsibility Experimental analysis. The analysis of variance per- formed on the Likert-type scales for self -attribution of responsibility produced a significant F ratio for the five experimental conditions (see Table 9) . The t tests demon- strated significant differences between means for global association (GA) and purposive commission (PC) , extended commission (EC) and PC, careless commission (CC) and PC, GA and justified commission (JC) , EC and JC, and CC and JC . All other comparisons were not significant (see Table 10) . These results indicate that subjects who received purposive commission attributions from their targets reported greater amounts of felt responsibility than all other experimental subjects except those receiving justified commission attributions. Also those subjects receiving justified commission messages reported greater amounts of 42 felt responsibility than all other experimental subjects except for those receiving purposive commission attributions. Table 9. Means and analysis of variance for self-attribution of responsibility scores Global Extended Careless Purposive Justified Association Commission Commission Commission Commission 3-57 2.94 3.36 4.12 4.07 N=14 N=18 N=16 N=24 N=ll Source of Variance df SS Between Within Total 4 16.35 4.09 3.01 <.05 78 81.14 1.30 82 96.49 Note: Self-attribution of responsibility was measured on a five-point Likert-type scale ranging from 1, strongly disagree, to 4, strongly agree. Table 10. t tests of experimental comparisons for self-attribution of responsibility scales Comparison GA.-EC 1.08 ns GA:CC .66 ns GA:PC 2.44 <.05 GA:JC 1.98 <.05 EC:CC .95 ns EC: PC 3.89 <,0 5 EC:JC 3.11 <.05 CC.-PC 1.77 <.05 CC:JC 1.70 <.05 PC:JC .45 ns Critical value for t (-.05, df=78) = 1.66. CHAPTER IV DISCUSSION While the role of personal responsibility in attitude- behavior discrepant situations has often been cited as a cause for attitude change (Brehm and Jones, 1970; Cooper, 1971; Collins and Hoyt, 1972; Cooper and Goethals, 1974; Reiss and Schlenker, 1977), the specific nature of the relationship between responsibility and attitude change has received limited scrutiny. The results of this study demonstrate that the attribution of responsibility when communicated to the actor by the recipient of the persua- sive message does have an impact on the actor's feeling of responsibility for the outcome of the action and the attitude of the actor toward the issue. ' Attributional messages from a target which hold that the actor behaved purposefully are given greater credence by the actor than other attributions of responsibility. In this study, subjects who received messages which contained purposive commission and justified commission attributions reported a greater magnitude of felt responsibility and greater attitude change than all other subjects. Although the hypotheses were not unequivocally sup- ported, the pattern of means was as predicted. Subjects 43 44 who received purposive commission attributions demon- strated significantly greater felt responsibility and attitude change than all other subjects except for those receiving justified commission attributions. While the means for felt responsibility and attitude change were greater for purposive commission subjects than for justified commission subjects, the effect size was not great enough to produce statistical significance with this small sample. It should be noted, however, that both messages contained all the elements which Heider (1958) maintains produce the maximum amount of responsibility attribution. The difference between the two levels lies in the addition of environmental justification included in the justified commission attribution. The justification manipulation in this study was a message stating that the recipient felt that the subject had to write the essay as a part of an assignment. It is plausible that the amount of justification perceived in the message manipula- tion in this study was simply insufficient to overcome the perception of intentionality of the subjects. Although level five was not significantly less than the fourth level, it is not valid to conclude that justification does not play an important role in subject's perception of personal responsibility and subsequent attitude change following counterattitudinal advocacy. Subjects may 45 indeed require more justification than was offered in this experiment. Foreseeabiiity. As stated in hypothesis three, subjects who received careless commission messages should have demonstrated more felt responsibility than subjects who received attribution messages not containing fore- seeabiiity manipulations (Heider, 1958; Shaw and Sulzer, 1964). Although the moans for careless commission subjects were larger than for subjects in the first two levels of attribution, the effect, size was not great enough to pro- duce statistical significance. The parallel prediction for attitude change, hypothesis three, also was not sup- ported although the means were in the predicted direction. According to Cooper (1971) foreseeabiiity is a necessary requisite to the self-attribution of respon- sibility and subsequent attitude change following counter- attitudinal behavior. "The analysis in terms of respon- sibility specifically precludes the arousal of dissonance in cases in which the consequences following from a free choice were unforeseeable. In such cases an individual can divest himself of responsibility and therefore experience no dissonance" (p. 355) . It is not clear from this analysis, however, whether foreseeabiiity of negative consequences is sufficient to produce attitude change in all circumstances. In the present study, the notion of 4 6 foreseeability was inherent in the experimental task. Subjects knew that their persuasive messages would be read by the students and certainly might have some impact. However, subjects did not report greater felt responsi- bility or attitude change unless they received attribu- tional messages which contained intentionality inductions. Foreseeability of negative consequences apparently was not enough to produce the necessary magnitude of felt responsibility to demand subsequent attitude change. In addition, even the messages that communicated attributions of foreseeability to the subjects were not sufficient to produce attitude change. While Cooper (1971) was able to demonstrate that subjects who could predict the con- sequences of their actions did demonstrate attitude change, it should be noted that the attitude measured was the subject's disposition toward his partner in the experiment. In the present study, the issue of dormitory visitation was unrelated to the target of the message. Although foreseeability was sufficient to produce attitude change in the Cooper study, it was not in the present experiment. The antecedent conditions may indeed be different in the two experimental situations. Also unresolved in these analyses is the question of whether or not the communication of purposive attribu- tions will produce attitude change in subjects who could not have foreseen the consequences of their actions. 47 Intentionality. The necessity of perceived inten- tionality was clearly demonstrated as a requisite for the self-attribution of responsibility and attitude change following counterattitudinal advocacy in this experiment. Goethals and Cooper (19 72) demonstrated that attitude change will occur in forced-compliance situations if the consequences of that action are aversive regardless of the intention of the subject. In the present study, it is doubtful that the subjects actually intended to produce attitude change in the recipients of their persuasive messages. More important was the apparent impression that the targets of the messages believed the subjects to be acting intentionally. It is plausible that subjects induced to perform counterattitudinal behaviors which may have negative consequences may actually hope that their attempts will be unsuccessful. Those who do indeed intend to produce a certain effect may see that result as less aversive than those who do not so intend. The accusation of intentionality by message recipients may increase the perceived aversivensss on the part of subjects who did not intend such results. The notion that the impression of intentionality as communicated to the subjects is more important than their actual intentions does not preclude a dissonance interpretation of the results. Consistent with 4 8 Festinger's (1957) claim that the greater the magnitude of dissonance, the greater the subsequent attitude change, increasing the subject's perception of personal respon- sibility and thus increasing his perception of the aversiveness of the consequences of the actions would cer- tainly increase the magnitude of dissonance experienced by the subject. Research on public commitment (Carlsmith, Collins, and Helmreich, 1966; Linder, Cooper, and Jones, 1967; Miller and McGraw, 1969; Miller and Burgoon, 1973) supports the idea that the individual ' s perception of his public impression is crucial to the dissonance formulation. Acts performed in private seldom produce dissonance-type attitude change in counLerattitudinal research. This is not to say that no internal processes (i.e. cognitive dissonance) are operating. Were this phenomenon strictly public behavior, it is likely that the attitude change demonstrated in this study might have been obtained with- out the parallel reports of personal responsibility. In addition to the attitude change observed in this experi- ment, subjects who chanqed their attitudes indeed reported more felt responsibility than other subjects, clearly an internal state. In addition to the knowledge that their behavior effected attitude change in the targets of their messages, subjects in this study also knew the impressions held 4 9 by those recipients of the degree of responsibility attributed to the subject. Research is needed which will directly compare attitude change and felt responsibility of subjects who know only of the consequences of their actions with those who also have knowledge of the respon- sibility attributions made by the targets of those messages This would help clarify whether feelings of personal responsibility for aversive consequences are due primarily to the circumstances of the action itself or the subject's perception of how that action is publicly viewed. There is research which indicates that the self -attribution of responsibility, as well as the self-attribution of other internal states, is indeed heavily influenced by the statements of others (Bom, 1967; Festinger, 1954; Rosenthal, 1964; Schacter and Singer, 1962; Nisbett and Schacter, 1966; Storms and Nisbett, 1970; Valins, 1966; Nunson and Kiesler, 197 1; Miller, Brickman, and Bolen, 19 75; Rosenbaum and Zimmerman, 195 9) . It is not simply a question of whether or not atti- tude change in forced-compliance situations is a result of the subject's impression management or some internal state but rather if the feelings of personal responsi- bility for the act and the consequences of that act are determined solely by the actor or by information which the actor obtains from others who are circumstantially 50 involved in the action, such as observers and those directly affected by the action. Festinger's social com- parison theory (1954) and researchers who focus on the self-attribution of internal states (Schacter and Singer, 1962; Nisbett and Schacter, 1966; Storms and Nisbett, 1970; Valins, 1966; VaJins and Ray, 1967) convincingly argue that such self-attributions are strongly influenced by environmental cues such as messages received by other persons in the immediate environment. A person who engages in a counteratti tudinal act which results in aversive consequences may or may not know whether or not he or she is responsible for those con- sequences until told so by those involved. It is plau- sible that only following the reception of such messages does the individual perceive himself as inconsistent and experience sufficient discomfort that he feels compelled to justify the act. If the attributional messages which that individual received are convincing and the meaning of those messages is inescapable for that individual ("You are responsible") then the pressure to readjust through changing his attitude to conform to the action is undoubtedly great. Personal responsib ility . Consistent with the Collins and Hoyt (1972) interpretation, this analysis supports a personal-responsibility-f or-consequences explanation of 51 the forced-compliance literature. Unique in this experi- ment is the strategy used to manipulate subject's feelings personal responsibility. The results of this study indicate that the communication of attributions of re- sponsibility from those affected by the subject's behavior will increase the subject's feeling of personal responsi- bility. Reiss and Schlenker (1977) have also demonstrated that the communication of attributions of responsibility by observers will have a similar effect. Collins and Hoyt ( 1972) also suggest that to induce persisting and generalizable attitude change in individuals one should "(1) encourage (if not demand) the individual to accept personal responsibility for his actions and (2) lead him to feel th.it his behavior in the group has important consequences" (p. 586). The present study, consistent with Munson and Kiesler (1974) and Reiss and Schlenker (1977) , suggests that one should not try to persuade an individual to accept responsibility for the consequences of an action but rather communicate to that individual a message that unequivocally attributes responsibility of consequences to that individual. It is also plausible that such messages will be maximally effective if the source actually observed the action (Reiss and Schlenker, 1^11) or was the recipient of the consequences of that action. 52 Cognitive dissonance. The results of this study are generally supportive of Reiss and Schlenker (1977) . They state that following counterattitudinal behavior, "People will engage in rationalization tactics only when all factors give the appearance of high responsibility for behaviors that produce aversive consequences" (p. 12) . In the present study all subjects received high choice and the appearance of aversive consequences but only those receiving the highest responsibility attributions actually reported attitude change. In agreement with Goethals and Cooper (1975) , Sogin and Pallak (1976) , and Reiss and Schlenker (197 7), it appears that it is possible to introduce postdecisiona 1 information which will increase the subject's feeling of personal responsibility and subsequent attitude change. "Thus, dissonance-type effects cannot be viewed as 'irreversible'" (p. 13). The postdecisional information in the Reiss and Schlenker study was designed to negate dissonance effects in certain circumstances, while the postdecisional informa- tion in the present study was intended to increase dis- sonance effects. It seems plausible, then, that dissonance outcomes may be manipulated either positively or negatively through the use of carefully constructed postdecisional messages. This is an important notion from a communication point of view. Cognitive dissonance, as well as other 5 3 explanations of counterattitudinal advocacy, have been primarily concerned only with attitude change effected by the counterattitudinal behavior and the circumstances under which it was performed. It is of interest to per- suasion researchers that attitude change may be predict- ably manipulated by creating the appearance of account- ability after the fact. Reiss and Schlenker (p. 14) also note that such results do not fully support an impression management explanation: "But the present results make it reasonable to conclude that subjects are managing their impressions for themselves as well as for any audience that is present" (p. 14) . It should be noted that in this study attitude change was accompanied by reported feelings of personal responsibility, a result which would be diffi- cult to explain solely i n terms of impression manage- ment. While not contradicting self-perception theory (Bern, 1967) , cognitive dissonance theory offers a more complete explanation of these results. By becoming an observer of one's own behavior, one may see inconsistencies and attempt to resolve them through several impression manage- ment devices, one of which may be attitude change. In the present study, subjects may have reported attitude change in order to maintain an appearance of consistency, but 54 why then would they need to report their feelings of personal responsibility? All subjects engaged in incon- sistent behavior, and the consequences of that behavior was made clear to them. However, only those held maximally responsible and hence reported greater felt responsibility demonstrated attitude change. If the perception of con- sistency were the primary motivator, it would seem that all subjects would have modified their attitudes to corre- spond with the behavior. Commensurate with the dissonance formulation, the notion that as the discomfort associated with negative consequences resulting from one's incon- sistent behavior increases, the pressure to avoid further discomfort increases proportionately, the threshold of tolerance for dissonance for the subjects in this experi- ment seemed to be reached only for those who were held maximally responsible for the outcome of their behavior. Unresolved, however, is the question of whether the greater magnitude of di ssonance was due to an increase in the feeling of responsibility for any negative outcome or if the increase in the attribution of responsibility increases the individual's perception of the aversiveness of the outcome. Limitations . The relatively small sample size, especially in the justified commission condition (n=ll) contributed to the lack of difference between purposive and justified commission. 55 With the semantic differential scales only 12 percent of the variance was accounted for by the justified commission vs. purposive commission comparison and only 11 percent with the Known Interval Scale. While the semantic differential scores produced a relatively high correlation (r=.75) with the Known Interval Scale, a higher correlation would have increased the probability that the two measures were both accurate indicators of the same phenomena. The lack of a greater correlation could account for the slight discrepancies in the mean patterns produced by the two measures. A weakness in the experimental messages allowed reinforcement to be inadvertantly confounded with the attribution of responsibility predictions. In addition to including intentiona 1 ity in the PC and JC messages, the messages portrayed positive effect on the part of the targets. The recipients stated, "I am glad that you tried to get me to change my ming . . . ," which could easily be interpreted by the subjects as a positive outcome . Future research. -\ replication of this study with the addition of a no-choice condition would help clarify whether or not the attribution of responsibility would create sufficient dissonance in subjects who did not feel as though they had a choice not to write a 56 counterattitudinal message. Typically, such subjects will not demonstrate dissonance-type attitude change following counterattitudinal advocacy. However, if attributions of purposive commission are introduced following counter- attitudinal advocacy, attitude change should follow only if impression management is a more powerful determinant of attitude change than dissonance in attitude-behavior discrepant situations. The magnitude of justification should also be examined more fully. By any standard, the justification provided to subjects in this study was minimal. Although being required to write a counterattitudinal essay may be seen as sufficient justification by some, it surely would not be by others. By providing increasing amounts of justification for the action, it would be possible to determine a threshold of justification in this experimental situation. APPENDIX I PRETEST QUESTIONNAIRE UNIVERSITY OF FLORIDA CENTER FOR COMMUNICATION RESEARCH One of the purposes of the Center for Communication Research is to assess public opinion on current issues. In order to help us gauge student opinion on several on- and off-campus topics, we ask that you please respond to the following topics by indicating your attitude on the scales that have been provided. Directions for the marking of the scales are below. PLEASE READ THE FOLLOWING VERY CAREFULLY On the following pages you will find a number of statements followed by several scales. Please mark each scale in the blank that BEST represents how you feel. For example, here is an item like the ones you will see: The United States should withdraw from the United Nations . Good :_ _: : : Bad Your job is to place a check mark (S) above the line that- best indicates your feeling toward the statement. For example, if you feel that U.S. withdrawal would be a very good idea, you would check as follows: Good X : _: : : Bad If you fell such a move (withdrawal) would be slightly beneficial, you would check as follows: Good : X : : : Bad If you feel neutral or indifferent about the proposition, or if you feel the scale is irrelevant to the proposition, you would check as follows: Good :_ : X : : Bad Remember : Never put more than one check make on a single scale and be sure that each check is in the middle of the line, not on the boundaries . 5 9 1. Dorms should have 24-hour visitation. pie asant o 1 i s h good uable unpleasant fo wise bad worth 1 val ess W >< w 1* a a P P « P >H Q E^ p >< P J p a rr. < O 2 H o P O £3 m p H W a U-l w P o < o w o p H p J CO H > CM Eh P u > U J J CO W 12 P D J a H P p P W p Q E-i W P U Eh S Eh X p H w H The practice of professors requiring books they have written as texts in their courses should be considered a conflict of interest. unpl easant wise plea: fool. ?ant LSh thl bad ess good wor valuable W >H w ■» a P p P & P X n Eh W >H P J i-l p < < o :;: KH O % n P « P H w P W PQ CQ o P Eh Cm O "> n P P CO CO 2 K D P 2 H W (X P w W Q Eh W P u Eh S3 H >: H w H The university of Florida should alter the grade point system to award a 4.5 to those students who show excellence in their class work. good valuable pleasant wise bad worthless unpleasant foolish W >H W >H P p p P P P >H Q Eh P >H « P P P Pn EH P u > o p P CO CO 2 P [3 P •z. H P P W w p P Eh w P u Eh P EH >', P 60 4. All lower division courses at U.F. should be taught by graduate teaching assistants. unpleasant wise pleasant foolish good valuable bad worthless W >H W >H Q Q pi J P^ Q X Q Eh W X « j W Oi < < o S H O Pi O 2 « W h w a w cq CQ o < o wo W H W J CO H > tii Eh W u > o W W LO co 2 a; D W 2 h w a; W W W Q En W 2 U Eh 2 H X 2 H W H 5. The U.F. libraries should be changed to a " closed stack" system, i.e. s tudents would request and receive books throm 5h library personnel . worthless valuable foolish pleasant wise unpleasant good bad W >< W >h Q Q Pi W Pi Q >h Q i-i w X « ^ J pi < < o £ H O Pi O 2 « w h w m um W o fifi o WO W H W J W H > ft Eh Cm O > O W W CO CO 2 Pi D W 2 H W (^ W w w Q Eh W 2 U Eh 2 Eh X 2 H W H 6. Traveling squat 3es f< Dr footbal . teams should be limited to 48 players. good unpleasant bad : ' pleasant wise foolish valuable worthless w x w X « ^ W Pi Q Q pcj ^ Pi Q X Q Eh W >h H w CQ w CQ oa o K o W W CO H W Pi W W 2 Q En W W W W 2 Eh 2 U Eh H X 2 W H 61 7. Students should be allowed to transfer at will with all credits being accepted between State University System schools. worthless_ wise unpleasant bad" valuable foolish pleasant good W >H W >H Q Q Pi i-l Pi Q >H Q Eh W >H w a ^ « < < o ,;] H O « O Z X J H W CQ W CQ m o Pi Pu Eh fci O > u J J CQ CQ 2 Pi D J z H W Pi W w w Q H H S3 u Eh S3 Eh :--: S3 H w H During this period of budget crunch, out-of-state students should not be admitted, to the University of Florida. va luable wort good fool plea iless bad wise ish unpleasant sant W >H W >H Q a cc; J Pi D Q Eh W >H W J J Pi <-.; < o r r; H O Pi U '.Z « hQ M W CQ W a a o c, < C w U :--i H W a CO H > a, Eh Pn o > (.'J J J CO 00 2 Pi 3 ^ S3 H W Pi w W LJ Q Eh U3 2 u Eh S3 Eh :.-: S3 H w H 9. In times of a budget crunch, faculty should teach heavier loads in lieu of doing research. pleasant_ fooli sh_ bad" worthless unpleasant wise good valuable W >H M >h Q Q PC .-1 PC Q >h Q Eh W >i « »J J Cr, J CO H > PL, Eh Cm rJ > o J >A CO CO S3 p: D a S3 H W Pi W W W Q Eh w S3 u Eh S3 Eh X S3 H w H 62 10. University College should be eliminated. wise valuable pleasant bad" follish worthless unpleasant good P >H N « J P H p n J en H CO is P h pq o; Q En w 2 Eh >h Q p < P P > p J ^ < •"■ Di CM bH P P 2 P X Q Eh W >H O Pi O 2 « P o P o p H W O > O p J CO p 53 p P u Eh 2 2 p H 11. All U.F. students should be required to attend at least one summer session to graduate. p leasan ba aluabl foolis t d unple good worth wise asant V e less h p >< a >H a Q p P P P >H n Eh W >H P P P a; < < o ,-' H o p n 2 « P H P a W P p o K < o w o p H P p CO H > Pa Eh P4 o > o p P CO CO 2 p P p 2 H W p W p H O Eh p 2 CJ Eh 2 Eh p 2 H w H 12 The Florida Twelfth Grade Placement Test should be abolished. foolish wise p leasant unple bad worth asant good aluable V less W >H w X p Q P P p Q n EH w X P P p p < CM Eh p e> > (T P p CO CO 2 OS D P 2 H P P W w P Q t^ w 2 u Eh 2 Eh p 2 M p H 63 13. U.F. should change from the quarter system to a semester system. worthless wise bad" unpleasant valuable foolish good pleasant w S* w >h a a « J « ^ M & < --: o < H W a W CQ ca o CU 1-5 CO H > a. H CO 2 a D H U W Q ^ u 2 2 E-i Q >h a Eh W >< O « O ;-: W J O u o ti H W a > o a hJ CO a 2 u w O EH X 2 uO H 14. Undergraduate studcMits in majors that require large monetary expenditures for their educations should be charged higher tuition. bad_ wise_ unpleasant valuable good foolish pleasant worthless W >H w >h Q Q g J « Q >i Q ^ W >H W J .J « O, ir< Ph o > o l-q iJ CO CO 2 rt D (J 2 H W C£ w w w Q Eh w 2 u E-h 2 Eh X 2 H w H 15. Registration period .:> should be set up according to the student's classi f ication , i.e., seniors enroll first, juniors, etc. good easant wise bad unp pl leasant f oo val lish wor thless uable W >H W >h Q Q « a Pi P >H n Eh cq >i W J J « < < Q & M O ^ C) 2 M J M LO CQ W CQ CQ O < O U o W h pq fj CO H > Cu Eh Dl, u > o ^ i-Q co CO 2 « D .J 2 H U OS U w W Q Eh u 2 CJ EH 2 Eh X 2 H w H 6 4 16. Physicians should be allowed to advertise their prices . foolish wise pleasant unpleasant good bad valuable worthless W i* H W iJ CO CO a H pq Q Eh 3 >n Q > >H Q OS O W O > u W >H « hi H W hJ CO z w Eh APPENDIX II EXPERIMENTAL MATERIALS UNIVERSITY OF FLORIDA GAINESVILLE May 25, 19 76 COMMUNICATION AND PUBLIC RELATIONS Department of Speech Dear U.F. Student, The University's Office of Communication and Public Relations is responsible for informing Florida high school students of the educational opportunities at the University of Florida. Part of the present recruitment program includes determining the responses of potential U.F. students to several proposed university policy changes. Rather than asking the students to respond to a policy state- ment in its technical form, we feel that a more realistic response can be assessed by having them say whether they agree or disagree with statements actually written by U.F. college students . In order to elicit the authentic feedback which we desire, we need your cooperation. Since we want al] viewpoints to be fairly represented, we are asking some students to write essays in favor of 24-hour visitation in dormitories and some students to write essays opposing the policy. Since we have already received the student paragraphs written in opposition to restricted visitation, we are asking that you write a paragraph in favor of this policy. Each high school stu lent will receive an envelope which contains letters from several U.F. college students. There will be some letters which are in favor of 24-hr. visitation and some which are opposed. In this way, the high school students can learn about both sides of the issue and decide for themselves where they stand. On the enclosed "memo" please write the most persuasive paragraph which you can think of in favor of restricted visitation . The members of the o) fice of Communication and Public Relations are grateful for your assistance in this project and we will let you know how prospective students feel about visitation policies just as soon as we can. Thank you again for your cooperation. Sincerely , Fred Kerlinger, Chrm. Office of Communication and Public Relations FK/wds 6 6 6 7 UNIVERSITY OF FLORIDA GAINESVILLE COMMUNICATION AND PUBLIC RELATIONS Department of Speech G8 Experimental Messages Global Association: Before I read the letters from the college students, I thought that 24-hr. visitation was a good thing, but I've changed my mind. Now I don't think it's such a good idea. I had the idea that it might be good before I saw the letters but now I ' m against i t . Extended Commission: I used to think that 24-hr. visita- tion was good, but I've changed my mind. Even though you couldn't know that your letter would make me change my mind and you didn't really mean to, it made me think that 24-hr. visitation isn't such a good idea. Before, I thought it might not be a bad idea, but not now. And even though you couldn't have had any idea that your letter would make me think different about it, it sure did. Careless Commission: Before I read your letter, I thought that 24-hr. visitation was a good thing. Even though you didn't really mean to change my mind, when you wrote the letter you probably knew that it might.' Now I think that 24-hr. visitation isn't such a good idea. Before, I had the idea that it might be a good idea, but now I don't think so. Maybe you didn't mean to make me think like you even if you knew that would happen, but anyway now I'm against it too. Purposive Commission: Before I read your letter, I thought 24-hr. visitation was a good thing. I am glad that you tried to get me to change my mind and that you knew your letter would convince me. Now I think that 24-hr. visita- i ion isn't such a good idea. Before, I thought it might be, but now I don't. I think you meant to make me I liink like you and knew that I would, but that's OK, you were right, because now I'm against it too. 6 9 Justified Commission: Before I read your letter, I thought that 24 hr. visitation was a good thing. Even though you had to write the letter as an assign- ment, I am glad that you tried to get me to change my mind and that you realized your letter would convince me. Now I think that 24-hr. visitation isn't such a good idea. Before, I thought it might be good, but not now. I think you wanted me to think like you and knew that would happen, but that's OK. I'm glad you had to do that assignment because now I'm against it too. 10 Opening and Closing Remarks for Experimental Messages Form I: Dear (subject's first name), I'm going to he a freshman at U.F. this fall. Since I'm going to live in a dorm, the letters were very useful to me. (experinv.Mital message inserted here) Thanks for helping me decide about dorm life. Sincerely , (bogus name) Form II: Dear (subject's first name), I'm really looking forward to coming to U.F. this fall. I will be living in a dorm my first year so the letters were really interesting. (experimental message inserted here) Thanks again for helping me make up my mind. Sincerely , (bogus name) Form III: Dear (subject's first name), • Since I'm going to be a college student this year and live in a dorm, the letters were really good. (experimental message inserted here) Now I know how I feel about visitation in dorms. Thanks. Sincerely , (bogus name) APPENDIX III POSTTEST QUESTIONNAIRE UNIVERSITY OF FLORIDA GAINESVILLE COMMUNICATION AND PUBLIC RELATIONS Department of Speech Dear U.F. Student, Thank you for your assistance in orientation of incoming freshmen. In order to complete our assessment of student attitudes about dormitory visitation policies, please take a moment to respond to the following items. Please mark each scale in the blank that BEST represents how you feel about visitation. Dorms should have 24-hour visitation. Pleasant Foolish Good Valuable Unpleasant Wise Bad Worthless 5h w Q J a < W PQ ca W H S ;::: E-i w w •r. H > oi hJ a Q Q Eh I* o Si H O O ^ ', vA o < o O W W a, E-i k, o o hH ■.'j 72 73 Please take a moment of your time to respond to the following items. For each item, make a circle around the number which BEST represents the degree of your agreement or disagreement with the statement above it. I feel responsible for changing the attitude of the high school student who responded to my letter. 2 o H I -i 2 o o >^ O oh h 2 E W to § ss E-< WO W 2 , 420-431. Burgoon, J., Burgoon , II., and Vaughn, D. The construc- tion of more precise and efficient rating scales. Unpubl i shed manusc rip t , lJru^vel:sTty~"of Florida"- T9 7 7 . Calder, B., Ross, J., and Insko, C. Attitude change and attitude attribution: Effects of incentive, choice, and consequences in the Festinger and Carlsmith paradigm. Journal of Personality and Social Psychology , 19 7 3, 25, 84- 99. 7 4 75 Carlsmith, J., Collins, B., and Helmreich, E. 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 of Personality and Social Psychology, 1966, 4, 1-13." Carlsmith, J., and Freedman, J. Bad decisions and dissonance: Nobody's perfect. In R. Abelson, E. Aronson, W. McGuire, T. Newcomb , 11. Rosenberg, and P. Tannenbaum (Eds.), Theories of cognitive consistency: A__s o urcebook . C h ic"agol Rand McNa 1 1 y 1968, p. 485-490. Cialdini, R. Attitudinal advocacy in the verbal conditioner. Journal of Personality and Social Psychology, 1971/17, 350-358. Cialdini, R. , Braver, 5., and Lewis, S. Attributional bias and the easily persuaded other. Journal of Personality and So cial Psycholog y , 19 1~A~, T6~^~6 3 1 - 637. Collins, B. , and Hoyt, M. Personal responsibi li ty- for-consequences : An integration and extension of the "forced compliance" literature. Journal of Experimental Socia 1 Psychology , 19 7 2, 8_ , 5 5 8 - 5 9 3 . Conolley, E., Wilhelmy, R. , and Gerard, H. The effect of pre- vs. post-ilecisional justification of subsequent task evaluations. Paper presented at the Western Psychological Association meeting, San Diego, 1968. Cooper, J. Personal responsibility and dissonance: The role of foreseen consequences. Journal of Personality and Social Psychology, 1971, 18, 354- 3 6 3. Cooper, J., and Goethals, G. Unforeseen events and the elimination of cognitive dissonance. Journal of Personality and S c > cial Psychology , 19 7 V, 23~, 44T- 44 5. Cooper, J., and Worchel, S. Role of undersired consequences in arousing cognitive dissonance. Journal of Persona lity and Social Psychology , 19 7 0, 1~6, 199-206. " 7 6 Eagly, A., and Whitehead, G. Effect of choice on receptivity to favorable and unfavorable evaluation of oneself. Journal of Personality and Social Psychology, 1972, ""22, 223-230~ Festinger, L. A theory of social comparison processes. Human Relations, 1954, 7, 117-140. Festinger, L. A_ theory of cognitive dissonance. Stanford, Calif.: Stanford University Press, 1957. Goethals, G. , and Cooper, J. Role of intention and postbehavioral consequence in the arousal of cognitive dissonance. Journal of Personality and Social Psychology, 1972", 23, 293-301. Goethals, G. , and Cooper, j. when dissonance is reduced: The timing of self-justificatory attitude change. Journal of Personality and Soical Psychology, 1975, 3_2, 361-367. Heider, F. The psychology of interpersonal relations. New York: Wiley, 1958. Holmes, J., and Strickland, L. Choice freedom and confirmation of incentive expectancy as determi- nants of attitude change. Journal of Personality and Social Psychology, 19 70", 14, SlPT^ Jones, E. , and Davis, K. From acts to dispositions: The attribution process in person perception. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 2). New York: Academic Press , 1965. Jones, R. , Linder, D., Kiesler, C, Zanna, M. , and Brehm, J. Internal states or external stimuli: Observers' s attitude judgments and the dissonance theory-self-persuasion controversy. Journal of Experimental Socia 1 Psychology , 19 68, £, 247-267 . Kelley, H. Attribution in social interaction. Morristown, N.J.: General Learning Press, 19TT~ Linder, D., Cooper, J., and Jones, E. Decision freedom as a determinant of the role of incentive magnitude in attitude change. Journal of Personality and Social Psychology^ 1967,6 , 245-254. 7 7 Miller, G., and Burgoon, M. New Techniques of persuasion. New York: Harper and Row, 1973 . Miller, G. , and McGraw , R. Justification and self persuasion of covin terattitudinal communication. Speech Monographs, 1969, 3_6, 443-451. Miller, R. , Brickman, P., and Bolen, D. Attribution versus persuasion as a means for modifying behavior. Journal of Personality and Social Psychology, 1975, 31, 430-441. Miller, R. , and Ross, M. Self-serving biases in the attribution of causality: Fact or fiction? Psychological Bui L etin , 1975, 82_ , 213-225. Munson, P., and Kiesler, C. The role of attributions by others in the acceptance of persuasive communications, Journal of Personality, 1974, 4_2, 453-466. Nel, E. , Helmreich, R. , and Aronson, E. Opinion change in the advocate as a function of the persuadibility of his audience: A clarification of the meaning of dissonance. Journal of Personality and Social Psychology, 196 9"," 12, 117-12 4. Nisbett, R. , and Schacbter, S. Cognitive manipulation of pain. Journal of Experimental Social Psychology, 1966, 2, 227-236/ Osgood, E. , Suci, G. , and Tannenbaum, P. The measurement of meaning. Urbaua, 111.: University of Illinois Press, 1957. Pallak, M. , Sogin, S., and Van Zante, A. Bad decisions: Effect of volition, locus of causality, and negative consequences on at. titude change. Journal of Personality and Social Psychology, 1974, 3_P_' 217- 227. Piaget, J. The moral judgment of the child. London: Paul, Trench, Trubner^ 1932 . Pittman, T. Attribution of arousal as a mediator in dissonance reduction. Journal of Experimental Social Psychology, 19 7 5T~TlT~" 5 3- 6TI Powell, F. Source credibility and behavioral compliance as determinants of attitude change. Journal of Personality and Social Psychology, 1965, 2, 669-676. Reiss, M. , and Schlenker, B. Attitude change and responsibility avoidance as modes of dilemma resolution in forced compliance settings. Journal of Personality and Social Psychology, 1977, 35, 21- 3 0 . — Rosenbaum, M. , and Zimmerman, M. The effect of external commitment of response to an attempt to change opinions. Public Opinion Quarterly, 1959, 23, 24 7-2 56. " — Rosenthal, R. On the social psychological experiment: The experimen tor's hypothesis as unintended determinant of experimental results. American Scientist, 1963, 51, 268-283. Rosenthal, R. Experimenter outcome-orientation and the results of the psychological experiment. Psycho- logical Bulletin, 1964, 61, 405-412. Schachter, S., and Singer, J. Cognitive, social, and physiological determinants of emotional state. Psychological Review, 1962, 69, 379-399. Shaw, M. , and Sulzer, J. An empirical test of Ueider's levels in attribution of responsibility. Journal of Abnormal and Social Psychology , 196 4, 6 9, 39-46 . Sherman, S. Attitudinal effects of unforseen consequences. Journal of Personality and Social Psychology, 1970," 16, 510-520. Smith, E. The power oi dissonance techniques to change attitudes. Public Opinion Quarterly, 1961, 25 625- 639. ~~ — Sogin, S., and Pallak, M. Bad decisions, responsibility, and attitude change: Effects of volition, foreseeability, and locus of causality of negative consequences. Journal of Personality and Social Psychology, 1976, 33, 300-306. ~ Storms, M. , and Nisbett, R. Insomnia and the attribution process. Journal of Personality and Social Psychology, 1970, f6, 319-328. Touhey, J. Retroactive inhibition of cognitive dissonance by increasing intensity of counter-advocacy. Human Relations, 1971, 26, 657-665. 7 9 Valins, S. Cognitive effects of false heart-rate feedback. Journa L of Personality and Social Psychology, 1966, "1, 400-408^ Valins, S., and Ray, A. Effects of cognitive desentization on avoidance behavior. Journal of Personality and S ocial Psychology , 19 6 7 , 77 345- 350. Zimbardo, P., Weisenburg, M. , Firestone, I., and Levy, B. Communicator effectiveness in producing public conformity and private attitude change. Journal of Personality, 1965, 33, 233-255. BIOGRAPHICAL SKETCH My father, Dewey R. Stinnett, retired from the Chesapeake and Ohio railroad after 52 years of service as a railroad engineer. My mother, born in Huntington, West Virginia, is a dedicated wife and mother. Although neither of my parents attended high school, they both highly valued education. Because of this, they sacrificed greatly to see their only son receive a quality education. However, in regard to the quantity of education, I don't believe they expected me to be in school until I was 32 years old. After high school graduation, I majored in speech and theatre at Marshall University in Huntington, West Virginia, where I received a Bachelor of Arts in 1968. The next year I entered the School of Interpersonal Communication at Ohio University and received a master's degree in 1971, following considerable turmoil both academically and personally. After a year of unemploy- ment and semi-employment, I received a position as instructor in the Department of Speech at the University of Nebraska at Omaha. Upon the urging of my friends, colleagues, and economic necessity, I decided to leave Omaha and pursue a Doctor of Philosophy degree. After one year in a doctoral program at West Virginia University I transferred to the University of Florida, where after much ill feeling as well as much good feeling I received the Doctor of Philosophy degree with a major in speech in August, 1977. Last year I was employed as an assistant professor in the Department of Communication and Theatre at Arizona State University in Tempe, Arizona, where I will be returning this fall. I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. ■V Michael Bur goon /Chairman Associate Professor of Speech I certify that I have read this study and that in my opinion it conforms Lo acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. ( '■ ; y A M C Judee K. Burgoon Assistant Professor of Speech I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. v/ / / / 'Thomas, J. Saine III Assistant Professor of Speech I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Barry ly. Sqhlenker Associate Professor of Psychology I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Marvin E. Shaw Professor of Psychology This dissertation was submitted to the Graduate Faculty of the Department of Spoech in the College of Arts and Sciences and to the Graduate Council, and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. August, 1977 Dean, Graduate School UNIVERSITY OF FLORIDA lllllilll 3 1262 08553 00.57