Participatory Design of a Hybrid Kinect Game to Promote Collaboration between Autistic Players and Their Peers

Abstract

We report on the participatory design of a hybrid two-player Kinect game to encourage complex emotion recognition and collaboration between autistic people and their peers. From its inception, autistic college students have been involved in an iterative process to design, evaluate, and redesign the game. The emotion recognition game has two playing phases. In the first phase, the players independently assemble pieces in a digital puzzle. In the second phase, players communicate in-person to agree on the appropriate emotion for the context and construct the emotional face for the body they have assembled together. We also designed collaborative reward games that require the two players to cooperate and one that encourages players to look at each other. In order to assess the level of in-game cooperation, we added a face tracking component that automatically quantifies collaboration and can replace time-consuming hand-coded evaluations. We report on how this game was designed and built by a team of autistic students, their peer mentors, a psychologist, computer scientists, and a graphic artist. Preliminary observations show that modeling of emotion recognition and collaboration by peers with stronger social skills is emerging as a central aspect of the effectiveness of our game. The participatory process has led to several design changes including one that dramatically increased player collaboration. We share insights and lessons learned that can guide others working in participatory design.