Instructional set, modeling and oppositional and general assertiveness : a multivariate analysis
The purpose of this study was the investigation of the effects of instructional sets and modeling on oppositional assertion while statistically controlling assertive predispositions by means of a multivariate analysis of covariance. Scores on the Rathus Assertiveness Schedule (RAS) were used as the covariate and ratings of two components of oppositional assertion, compliance content and global assertiveness, were the dependent variables. Additional measures were taken using the whole IRAS scores and the Poland Factor 1 (PF'1) "Oppositional Assertion" scores of the RAS in a repeated measures ANOVA design.All subjects in this study were female undergraduate students of Ball State University. The treatment of the, subjects was as follows: The entire potential subject pool (students taking an introductory general psychology course in the Department of General and Experimental Psychology, Ball State University, Spring Quarter, 1976) was administered the RAS being given no information regarding the purpose of this task. The RAS protocols of the male students were removed while those from the female students were retained. Three weeks later a female confederate recruited female subjects for "an experiment on women." These subjects were then randomly assigned to one of the three treatment conditions, videotaped assertive modeling alone, modeling with behavioral instructions focusing on the known components of assertiveness, or modeling with instructions focusing on the psychoanalytic constructs of ego autonomy. A no modeling control treatment could not be included because of subject and design problems. The effects of pure modeling effects can be inferred by use of the PF 1 of the RAS. Each subject was individually shown the Training Film and immediately afterwards was given a series of ten videotaped test situations to which she was to respond. These responses were recorded on audiotape andgiven to three judges for rating. Subjects were then asked to complete a RAS and a general nonempirical questionnaire, and were then debriefed by the experimenter.Three raters were trained by the experimenter and each heard and evaluated every tape. Three-way interjudge reliability was very high with r = .882 for compliance and r = .798 for global assertion. These ratings were summed and entered in the MANCOVA design.Results of the data analysis revealed surprising outcomes. The RAS did not covary significantly with the dependent behavioral variables. With the possibility of covariate analysis obviated; a MANOVA was performed using the remaining behavioral data. No F2 value exceeded 1.00 indicating no behavioral difference between the treatments. There was also no significant difference between the pre- and post-test scores of the full RAS. Significant results were obtained using PF 1 of the RAS. The treatment incorporating psychoanalytic constructs in the instructions showed the greatest magnitude of increased oppositional assertion (t = 5.38, p_<.001).' The group with behavioral instructions also increased significantly (t = 2.77, P_<-05) while the group receiving modeling alone did not significantly improve (t = 0.06, P_>.05), Additional post hoc analysis indicated that the randomization had failed to control. all variables. The behavioral group was significantly different from the modeling-only group (C = 1.76, p_ .001) at pre-test, with no significant differences in the other comparisons. On post-test both the behavioral and the psychoanalytic groups here significantly (t = 6.96, P <.001; t =7.03, p <.001) different from the modeling only group but not from each other.The conclusions were made that the multivariate approaches had failed to reveal significant changes because of severe (>33%) subject attrition (N after attrition = 39). The significant findings using PF 1 of the RAS were seen as evidence that the treatments had been effective in modifying assertion, with the treatment including psychoanalytic constructs perhaps being more effective. No generalization to other forms of assertiveness was seen.The recommendation was made for additional research on inter-theoretical application of assertion training. The possibility of salutory therapeutic effects resulting from combining behavioral and intra-psychic insight approaches was suggested.