The Empirical Nature of Behavior Therapies

Abstract

This paper presents an analysis of the conceptual and methodological practices of behavior therapies. Our analysis focuses on the common practice of identifying specific organismic events and empirically evaluating the effects of well-specified independent variables on the probability of those events. The practice is most commonly identified as the experimental analysis of behavior (Skinner, 1970). We present this analysis because it appears that many behavior therapists have either rejected the study of private events or, in their attempts to incorporate them, have abandoned or ignored the experimental analysis of behavior and have substituted ambiguous concepts and mentalistic explana tory practices. It therefore seems appropriate to demonstrate that any psychological problem can be construed in terms of specific organismic events, independent variables which are said to affect those events, and empirical evidence that the independent variables do affect the behavior in question. This general framework has critical implica tions for current conceptual practices and for the way in which behavior therapy achieves rapprochement with other orientations.