This initial evaluation has limitations, in particular the small number of participants and sessions. The regression analyses are based on a small number of observations and are, therefore are not conclusive. However, the scale of the study is appropriate based on where we are in the development of the intervention. From 2002 to 2004, the National Institutes of Health (NIH) sponsored a working group to discuss the methodological challenges of conducting research on interventions for children with ASD [Lord et al. 2005; Smith et al. 2007]. The results of this working group included two papers. One paper discusses the challenges inherent in conducting intervention studies with children with ASD such as participant recruitment, appropriate outcome measures, and putting research interventions into practice [Lord et al. 2005]. The second paper outlines four stages for conducting intervention research: (1) demonstrating efficacy; (2) compiling and pilot-testing an intervention manual; (3) conducting randomized clinical trials (RCTs) that evaluate the intervention under controlled conditions; and (4) conducting studies that evaluate intervention use in community therapeutic practices. Since the AVP is a new technique for working with this population, the first phase of research, which aims to provide a “proof of concept” [Smith et al. 2007] of the technique, is appropriate. While working with small sample sizes at this phase of research limits generalization of results, using small sample sizes addresses some of the challenges of this phase, such as limited resources or procedures that require adjustments and are not standardized. So while future work by clinical researchers can develop and evaluate a standardized intervention program, the current research has taken the first step by identifying a new intervention technique incorporating AVPs and offers evidence of its efficacy on reciprocal social interaction with peers.