Our previous work (IMFAR 2009) introduces Authorable Virtual Peers (AVPs), a technology tool designed to scaffold contingency and reciprocity skills during peer social interactions. An AVP is a life-sized, child-like, animated character that interacts with children in three modes: children can collaborate face-to-face with the virtual peer, using speech and gestures, to create a story; they can create new stories for the virtual peer to tell with other users; and they can select pre-planned responses for the virtual peer while it interacts with another person. In this study, we developed and evaluated a comprehensive group intervention program incorporating the AVP.
Objectives:
We asked, (1) can an AVP be integrated into a group intervention program? and (2) Do AVP interactions help children with ASD engage in reciprocal social interactions with their peers?
Methods:
Working with two therapists at a developmental disorders clinic, we created a 10-week program that incorporated the AVP as a component of a social group curriculum (Baker, 2001). The program included didactic instruction on specific social skills, group activities and dyad role-play. Each week, four of the children used the AVP system prior to the role-play task. They create new story components for the virtual peer, and then took turns with a partner controlling and interacting with the virtual peer. The analysis examined appropriate use of reciprocity skills during the role-play task. We used standard least-squares regression to see if using the AVP, the number of weeks in the intervention or language ability, along with interactions, were significant predictors of appropriate reciprocity. Eight children, ages 8-12, participated (7 included in the analysis).
Results:
We found a significant effect due to the AVP (p<.003) that suggests children were more likely to use appropriate reciprocity during the dyad role-play if they first interacted with the AVP. We found a significant effect due to week (p<.002) that suggests children's use of appropriate reciprocity increased over the course of the intervention program. We also found a significant effect due to the interaction of language and week (p<.01), suggesting language ability negatively impacted the increased use of appropriate reciprocity.
Conclusions:
Thus children were more likely to use appropriate reciprocity skills after using the AVP, and their appropriate use of reciprocity increased over the course of the intervention. In this technology demonstration, we will present the software, results of our study, and video examples of children engaging in AVP interactions.