Therapist personality and treatment outcome: a test of the interaction hypothesis using the Campbell A-B scale

The A-B scale was developed to distinguish therapists with a high rate of therapeutic improvement with schizophrenic patients from those with a low rate. The present study tests the hypothesis, developed by Whitehorn and Betz, expanded by McNair and Lorr, of an interaction between a therapist factor, i.e. the A-B score (identified in this study by the Campbell et al. revision of the Whitehorn and Betz A-B scale), patient diagnosis, and treatment outcome. An automated data system identified all of the adult patients treated by 42 therapists during their training at a psychiatric center containing inpatient and outpatient facilities. The patient group was broken down according to diagnosis and whether or not psychotropic drugs were a major treatment modality. Using a linear logistic regression model, the slope of the line specifying log odds of improvement in relation to therapist A-B was found to be significantly different for schizophrenics and neurotics treated without drugs. (No relationship to A-B was found for drug-treated patients.) For those therapists who had treated both neurotic and schizophrenic patients (N = 19) without drugs, A-B was found to be significantly associated (p = .075) with a measure of dependence between patient type and outcome. These two findings were consistent with the hypothesized relationship between the A-B dimension, patient diagnosis, and case outcome.