| In the past, psychotherapy outcome research has been directed toward detecting treatment effects, mostly ignoring other sources of variability, particularly patient and therapist factors. However, in recent years, leading researchers have become interested in therapist effects, which are defined as the degree to which therapists vary in the outcomes they produce, apart from the effects due to treatment. Specific ingredient models of psychotherapy tend to focus on the benefits that result from particular treatments whereas common factor or contextual models of psychotherapy give primacy to the person of the therapist. Consequently, if the ratio of variability due to therapists to variability due to treatments is large, evidence accrues for a common factor or contextual model of psychotherapy.;The present study estimated the size of therapist effects as well as treatment effects in a particularly well conducted large-scale psychotherapy outcome study: the National Institute of Mental Health Treatment of Depression Collaborative Research Program (NIMH TDCRP). The estimate of therapist effects under cognitive behavioral therapy (CBT) and interpersonal psychotherapy (IPT) conditions was 12.4 percent of outcome variance in the termination score analysis, whereas the estimate of treatment effects was zero percent. On the other hand, when change rate across time was modeled, 19.7 percent of within-treatment variance in the rate of change was identified with therapist effects, and no more than seven percent of variance was ascribed to treatment. These results, indicating that therapist effects were greater than treatment effects, confirm the prediction of the common factor approach. The implications of the result for practice and future research are discussed. |
