APPLYING STATISTICAL CONTROL-THEORY TO BRING TOGETHER CLINICAL SUPERVISION AND PSYCHOTHERAPY-RESEARCH

A three-factor experiment was designed to evaluate the potential for improving first-session performance of 6 trainee-therapists as measured by postsession client ratings N = 48. Rather than through standardizing therapist behavior by using a training manual beforehand, therapist effects were handled through statistical control. This method not only provided control so that findings were not confounded with some kinds of therapist effect but also identified which therapists needed supervision to bring their performance more in line with others in the group. Reducing the number of therapists requiring "standardization" should lower training costs and enhance individual freedom. In the present study, more frequent empathic responses led to statistically significant beneficial effects on five of seven outcome measures. Suggestions volunteered by therapists improved overall session ratings on three variables and increased scores on a measure of passive dependence. Whether or not therapists asked questions had no main effects on any tested outcome measure.