ABSTRACT
This chapter shows that the research strategy of employing a moderate degree of subject heterogeneity in outcome investigations, and subsequently using regression techniques to isolate sources of variance attributable to subject characteristics, may have considerable potential for determining the optimal matching of clients with treatments in future research. The investigation combined progressive relaxation training and cognitive restructuring into one treatment package. Considerable attention within the psychological literature has been devoted to anxiety as a clinical problem. Male and female respondents to an ad for anxiety-reduction therapy were randomly assigned to transcendental meditation, behavior therapy, self-relaxation, or a wait list control group. Clients who reported perceiving more internal locus of control benefitted more from transcendental meditation than clients who reported greater external locus of control. The relationship of client characteristics to outcome was evaluated separately for each of the treatments via stepwise multiple regression analyses.
