External validity of role play assessment of assertive behavior

Forty female subjects, equally divided into high- and low-assertive groups on the basis of self-report data, participated in an assertion-eliciting situation. Half of the subjects were advised that the situation was not real and asked to role play their responses. Half of the subjects were uninformed that the situation was staged (natural behavior condition). High-assertive and low-assertive subjects were significantly different in both their self-report and their behavioral performance. Role play and natural behavior conditions did not differ, nor was the overall multivariate test of interactions significant. The findings indicate that role play assessment was comparable to naturally occurring behavior and provide support for the continued use of role play assessment in social skills research.