This investigation compared the relative effectiveness of two social skill training packages for patients diagnosed either schizophrenic or non-psychotic. Patients who scored low on a global measure of social skill were randomly assigned to one of three training conditions: social skills training with modeling, social skills training without modeling, and a behavior rehearsal control. Effects of training were assessed on brief role playing and extended interaction tasks. The results indicated that both social skills training programs had positive effects. However, modeling was essential in improving the performance of schizophrenics, but was unnecessary for non-psychotics. The effects of training transferred to semistructured inter-personal conversations. The importance of designing social skill training programs for relatively homogeneous populations was discussed.

No abstract

The effect of response class on self-monitoring accuracy was investigated in a clinical population. 16 in-patients of a VA hospital self-monitored verbal and nonverbal behaviors in a couterbalanced design. The results supported suggestions of superior self-recording accuracy with nonverbal target behaviors. The implications for clinical use of self-monitoring in assessment are discussed.