Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)




The present study was designed to demonstrate the possible applicability of the concepts and findings from attitude change research to a counseling or psychotherapy situation. Expanding and elaborating on a study by Bergin (1963), the effects of source credibility and setting credibility were investigated with respect to their independent and combined effects upon changes in the self- referring attitude of masculinity following a male subject's reception of communication indicating a discrepancy between his own and the communicator's judgment. Changes in anxiety as a result of receiving the discrepant communication were also measured.

Each of 60 male college student volunteers was assigned randomly to one of the four experimental conditions created by combining high versus low communicator credibility with high versus low setting credibility. An additional 30 subjects served as a control group. All subjects were pre-tested on a "Self-Analysis Rating Form", the Spielberger Anxiety Scale and the Edwards Personal Preference Schedule. In a subsequent interview with the "expert", each experimental subject received communications indicating essential agreement between himself and the communicator on all his self-ratings except masculinity, for which there was a wide discrepancy. Following the interview, each subject again filled out the self-rating form and the anxiety scale. In addition, subjects were asked to supply information concerning their reactions to the "expert" and to the experiment. Control subjects, who did not have a feedback interview, also filled out the self-rating form and anxiety scale a second time. Changes in self-rated masculinity and in anxiety were assessed by analyses of covariance and follow-up procedures as deemed relevant and appropriate.

The results indicate that changes in self-rated masculinity in the direction of the communicated discrepancy were significantly greater for both the high communicator and the high setting credibility subjects than for their respective low communicator and low setting credibility counterparts. No significant interaction effects were found. Relatively greater changes were found for high communicator than for high setting credibility. When the direction of change advocated by the discrepant communication was taken into account, it was found that subjects in both the high communicator-high setting and high communica tor-low setting conditions tended to change regardless of the direction advocated. Subjects in the low communicator low setting condition tended not to change regardless of the direction advocated. Subjects in the low communicator high setting condition tended to change if the direction advocated was toward higher masculinity ratings, but not to change if the direction advocated was toward lower masculinity ratings.

Smaller changes in anxiety were found in the condition yielding the greatest mean amount of change in masculinity ratings. >In addition, correlations between changes in masculinity ratings and changes in anxiety, with the direction of the change advocated taken into account, reveal the following: If the direction of change advocated is toward higher masculinity ratings, there is no significant correlation between changed ratings and anxiety. If the direction of the change advocated is toward lower masculinity ratings, there is a significant negative correlation between changed ratings and anxiety. An examination of the latter relationship suggests several highly tentative interpretations. First, subjects who change their self-ratings either are not emotionally aroused or have sufficiently worked through the discrepancy by the time of the second testing. Secondly, subjects who do not change their self-ratings remain anxious when retested. A continuously monitored state anxiety level would be needed in order to shed some light on the relationships between attitude change and emotional arousal.

Results of the study reaffirm the importance of credibility in inducing attitude change and indicate further that the source of the communication and the setting in which it occurs, contribute independently to its credibility.

High levels of credibility appear to induce changes in the belief component of an attitude while low levels or no credibility do not. However, a moderate level of credibility appears to induce change only if the advocated change is in a direction we may assume that the subject considers favorable. Thus, some credibility is necessary to induce change even when the subject regards the change as desireable.

These conclusions, most of which are highly tentative, have implications for both attitude change research and clinical practice which should be followed up in several different research studies.