Revising Reciprocity: Technology Tools for Creating Social Interactions

Background: The difficulties children with high-functioning autism (HFA) have engaging in reciprocal social interactions with their peers affect relationships, education, and employment opportunities. Reciprocal social interactions rely on contingency – maintaining a conversation such that what one says follows from what was previously said – to manage the topic of conversation and engage the other person. Previous studies suggest that children with HFA are less likely to be contingent compared to other children. We propose technology tools as a scaffold for the development of contingency and reciprocity skills.

Objectives: Develop technology tools that allow children with autism to plan, engage in, observe, and revise social interactions. Examine contingency of the social interactions across revisions.

Methods: Design-based research study that used iterative design sessions with individual children to develop tools for planning, engaging in, observing, and revising social interactions. Seven children with high-functioning autism (1 girl), ages 8-12 participated. Severity of autistic social impairment was assessed using the Social Responsiveness Scale (SRS); T-scores were all clinically significant and ranged from mild/moderate to severe. Expressive and receptive language ability was assessed using the Peabody Picture Vocabulary Test (PPVT-4) and the Expressive Vocabulary Test (EVT-2); age equivalents of language ability ranged from 6:2 to 15:9. All participants were enrolled in social groups at a local autism clinical program.

Results: We modified our virtual peer to be authorable, and to function in three modes: collaborate, operate and author, and this study provides evidence of children’s use of contingency with the authorable virtual peer. For the collaborate mode, we adapted the “Create a Story” press from the ADOS as a collaborative narrative task where the child and virtual peer work together to tell a story using five objects. This task was developed to (1) be age appropriate; (2) require turn-taking as well as obtaining and incorporating peer input; and (3) use physical objects to enable embodied play to create a connection between the child’s physical world and the virtual world of the virtual peer. In operate mode, the child uses an interface to select pre-recorded and animated utterances for the virtual peer while it collaborates on the same task with another person. In author mode, children create new utterances for the virtual peer and design their own interface for operating the virtual peer. Evidence from the design study suggests Authorable Virtual Peers can: (1) be used to identify behaviors that may be affecting reciprocity; (2) provide a scaffold which allows children to employ contingency skills related to reciprocity that are difficult for them in conversation with other children, including: adding new information, asking questions, listening, and providing feedback; and (3) enables children to monitor and modify social behavior.

Conclusions: This technology demo will demonstrate how tools for authoring and operating virtual peers enable children to employ contingency skills related to reciprocity while controlling, monitoring and modifying the social behavior of a virtual peer. We will also demonstrate how the system can be adapted to support different abilities.