Effects of modeling on self-disclosure

The research was an experimental investigation of the differential effects of various modeling conditions upon self-disclosing behavior. One purpose of the study was to determine whether or not modeling treatments can significantly influence self-disclosure. The second purpose was to evaluate what differences exist in self-disclosure among males and females. Two philosophical assumptions were implicit in the research. The first assumption was that self-disclosure is an essential ingredient for interpersonal adjustment. The second was that self-disclosure is a construct that can be operationally defined and measured. Both assumptions were supported with empirical research.Self-disclosure was defined as an interpersonal process in which a person communicates intimate information and personal feelings or actions to another person. The disclosed information may be past or present and involves a personal risk when the discloser reveals his or her private subjective experiences. The elements of self-disclosure were delineated and a theoretical framework for analyzing the parameters of self-disclosure was discussed. Four measurements were considered essential for obtaining an accurate assessment of self-disclosure. The depth ofdisclosure was evaluated by trained judges who rated audiotapes according to a Revealingness Scale. Duration was assessed by timing the length of each subject disclosure unit. The third measure was the difference on a pre-post "willingness to discuss" instrument. Finallya self-report instrument indicated the subjects' evaluation of their personal disclosures and their partners' disclosures.The population consisted of volunteer undergraduate students enrolled in Introductory Psychology courses at Ball State University. A stratified random sampling procedure was applied to select the subjects from among the volunteers. The subjects were randomly assigned to one of four treatment groups. Each group had eight males and 16 females yielding a total of 96 subjects. The four groups consisted of an intimate in vivo modeling condition, an intimate videotape modeling treatment, a superficial videotape modeling treatment and a control. Each videotape modeling condition consisted of two persons taking turns sharing with each other on intimate topics for about 18 minutes. The in vivo condition involved a model sharing with a subject intimately for about 15-18 minutes. Subjects were exposed to a modeling condition and then requested to pair with another subject and disclose personal information in a structured setting.Two multivariant null hypotheses were tested. The multivariant hypotheses included several sub hypotheses related to each dependent measure of disclosure. The first hypothesis stated that no significant difference in the overall F would exist between the vector of means for males and females. The second hypothesis stated that no significant difference in the overall F would exist between the vector of group means.Multivariate statistical analyses were applied to test the null hypotheses. No significant differences were found among the four groups on any of the disclosure parameters. Likewise, males and females did not differ in self-disclosure. Several a priori hypotheses were analyzed to compare the effectiveness of the modeling conditions. Treatment conditions did not have a differential effect on self-disclosure.Although the overall F was not significant, a significant difference was located through univariate analysis of the sub hypotheses. Male subjects differed from females on the pre-post "willingness to discuss" instrument. Females greatly increased their scores on the posttest, but the scores of the males decreased on the posttest. The difference was significant at the .05 level.The data was further analyzed to explore the relationship between religious commitment and self-disclosure. Three categories of religious commitment were included. Multivariate statistical analysis revealed that subjects indicating strong religious commitment were significantly different on the pre-post measure than subjects in the no religious commitment classification. Deeply religious students increased on the posttest scores while students with no religious commitment decreased. Subjects in the latter category initially perceived themselves as willing to disclose much personal information, but decreased after experiencing the dyadic interactions. The difference was significant at the .025 level.Recommendations and evaluation of the research findings were discussed. Considerations for further research on facilitating self-disclosure through dyadic models were delineated.Thesis (D. Ed.