Objectives: To examine viewing patterns of young adults with ASDs participating in social interactions during a community art recreational program.
Methods: Participants in this study were 6 (4M, 2F) young adults 18-21 years old with higher-functioning ASDs. Data were collected during the Artism program, which consisted of weekly 90 min sessions that allowed social opportunities for the participants as they completed art-related activities. Participants all wore eye-trackers concurrently in a 30 minute pilot session aimed at testing the feasibility and potential of these systems. This abstract presents preliminary findings of 4 of these participants (2M, 2F) involved in a single session, using interval coding from the scene camera. During this session, participants were paired in dyads to complete an activity. By collecting data on the participants simultaneously, we were able to observe the participants engaging in within dyad peer-to-peer interactions, across dyad peer-to-peer interactions, and peer-to-instructor interactions.
Results: Participants largely worked with their partners independently of other pairs. We examined periods in which participants were involved in conversation with one another or the instructors, operationally defined as being either the target of someone speaking or speaking to someone else. Participants spent half their time in conversations (M=49%, SD=9%). They spent one third of this time looking at their conversational partners centered in their field of view (M=29%, SD=13%), looking at them peripherally (M=33%, SD=14%), or not looking at their partner (M=38%, SD=7%). Participants spent a small time looking other others while not engaged in conversation (M=15%, SD=10%), and overall spent a third of their time looking at others (M=31%, SD=8%). We additionally examined a brief lecture episode in which the instructor provided instruction regarding the upcoming project. Though there was a large amount of variability in participant’s behavior, the correlation between participants’ proportion of centered looking at either conversational partners or the instructor in the lecture episode was r=.76. This was not significant (p=.24) due to the small number of participants.
Conclusions: Though highly preliminary, these results suggest that low-cost, “home built”, head-mounted eye-tracking systems are a viable technology for use with young adults with ASD in relatively unconstrained natural interactions and in community settings. Further work will refine our technology and investigate the utility of these devices for observational and experimental paradigms as well as their potential for tracking the effects of interventions.