Group conversational-skills training with inpatient children and adolescents. Social validation, generalization, and maintenance.

Two groups of inpatient youths (n = 4 and n = 5) received group conversational-skills training. Conversational component behaviors of both groups at baseline were compared to the behaviors of normal youth from the community. For each inpatient group, component behaviors on which they were most discrepant from normal youths were selected for training. Multiple-baseline designs across component behaviors were used and patients were trained to socially-validated criterion levels. Effectiveness of training was demonstrated through (a) behavioral ratings during dyadic, unstructured conversations with other members of their group, (b) behavioral ratings during conversations with unfamiliar nonpsychiatric persons, (c) global ratings of conversational effectiveness by nonpatient peers, (d) generalization to in vivo conversations, and (e) maintenance of improvement at 1- and 3-month follow-up assessments. Results were replicated across both groups and indicated that training effects were socially validated, generalized to conversations with unfamiliar persons and to in vivo conversations, and maintained over time.