Contingent Reinforcement Modeled Consequences

Abstract Children with autism who have deficits in social and communication skills are often limited in their capacity to learn play skills. Video modeling has been used to teach scripted play skills to children with autism and many variations have been used to determine which techniques are most important. Children have been observed to readily imitate behaviors exhibited by a model in the presence of a model and have generalized those responses to a new setting in the absence of a model. The purpose of this study was to determine whether observing consequences provided to the model (positive reinforcement or no observed positive reinforcement) influenced skill acquisition of video modeled play skills. An A-B-A-C-A-X experimental design counterbalanced for order of presentation, and X representing the most effective treatment condition was utilized with four children diagnosed with autism. The results of the study indicated that neither contingent reinforcement (CR) nor no reinforcement (NR) conditions were more effective than the other for teaching modeled actions and scripted verbalizations during play. Decreases in unscripted verbalizations were observed during both CR and NR video modeling conditions