Effects of Video-Based Group Instruction On Generalized Social Behavior of Adolescents with Severe Autism

Background:  Numerous strategies for teaching social skills to individuals with autism have emerged in the intervention literature. Despite some practices with efficacy for teaching targeted behaviors, there is minimal evidence documenting generalization of social behavior to novel settings following social skills interventions for individuals with autism. Plavnick, Sam, Hume, & Odom (in press) recently piloted and assessed a version of a social skills group procedure that included video modeling and several strategies known to promote generalization including multiple exemplar training, programming common stimuli, and general case programming (Stokes & Baer, 1977). The intervention, video-based group instruction (VGI), led to the acquisition of complex social behavior and parents reported the generalization of targeted skills, though generalization was not assessed through direct observation of behavior. Additional research focused on generalization of skills following exposure to VGI is needed to better understand the efficacy of this procedure.

Objectives:  The purpose of the present investigation was to evaluate the extent to which social initiations and responses taught during VGI sessions generalize to novel settings. A secondary purpose was to identify additional components, if necessary, that mediate generalization of social behavior following VGI.

Methods: A single-case experimental research design was used to identify a functional relation between VGI and generalization of social behavior by 11 adolescents diagnosed with autism and a severe cognitive impairment.  A multiple baseline across groups design was used to evaluate the efficacy of VGI to promote generalization of social initiations and responses.  Participants were observed in multiple settings across a public school prior to and following implementation of VGI, which was administered only in the classroom environment, to observe instances of generalization.

Results: Preliminary data suggests some generalization of social responding, though variability across participants has been observed. Individuals with more severe deficits in language and social behavior prior to the intervention were less likely to engage in generalized social behavior following the intervention. Additional strategies, including training across multiple environments and peer mediated interventions, were needed to facilitate generalization of social behavior for all participants. Educational service providers and parents reported increased instances of social behavior, suggesting that VGI may promote generalization of social skills for individuals with autism.

Conclusions: The results of the study show that VGI may be a strategy capable of teaching social behavior to adolescents with autism and promote generalization of social skills in untrained settings. The findings extend previous research in several ways. First, the present experiment was designed to explicitly assess generalization of social behavior rather than evaluating generalization as a secondary purpose. Second, the intervention involved adolescents diagnosed with autism and who demonstrated a severe cognitive impairment. Minimal research has included such participants to date. Finally, the intervention was delivered in a public school environment and incorporated into the daily curriculum of participants. These extensions have several implications for future researchers and practitioners that will be discussed.