One assumption of the Role Category Questionnaire (RCQ) as a measure of cognitive complexity is that the sample of constructs it elicits represents a stable feature of individual difference. Two experiments investigating the RCQ's stability were conducted. Results for Study 1 demonstrated that changing RCQ instructions elicits a significantly greater amount of constructs compared to previous studies. Study 2 also showed varying instructions for construct, production increased mean number of constructs generated and changed the rank‐ordered relationships of complexity scores. Results suggest the RCQ's method of sampling the cognitive system yields unstable and sample‐bound scores. Possible refinements in assessing cognitive complexity are discussed.
Notes
Mike Allen is an Assistant Professor, and Edward A. Mabry is an Associate Professor, in the Department of Communication, University of Wisconsin‐Milwaukee; Mary Banski is a doctoral student at the University of Texas at Austin. Mary Banski, Marjorie Stoneman and Paul Carter were M.A. candidates in Communication at UW‐Milwaukee when this report was prepared.