Early Intervention for Socially Withdrawn Children

Abstract

Eighteen studies were reviewed in which single-subject methodology was employed to evaluate treatment of social withdrawal in preschool children. Studies were systematically analyzed for a number of variables, including description of target subjects, type of intervention, length and intensity of intervention, primary intervenor, and several outcome variables, including percent of nonoverlapping data points from baseline to intervention phases and an overall outcome rating based on a 3-point scale. Studies were analyzed for covariation between outcomes and study characteristics. Results indicated that (a) interventions in which target subject interactive behaviors were directly reinforced and measured were the most successful interventions, (b) target subjects' unreinforced responses to trained peer confederates exceeded that of target subjects' unreinforced social initiations to confederate peers, (c) reinforced confederate peer social interactions exceeded unreinforced interactions, and (d) unreinforced modeling had little effect on subsequent social behavior. Furthermore, setting and skill generalization generally did not occur, maintenance was rarely assessed, and information about the characteristics of subjects included in such analyses was extremely limited. Implications for further research are given.