24141
Facetime Vs. Screentime: Decreased Modulation of Gaze Patterns to Live and Recorded Social Stimuli in ASD
One of the primary characteristics of individuals with ASD is atypical eye gaze during social interactions (Kanner, 1943), but most quantitative data on how individuals with ASD explore social stimuli is based on eyetracking studies of participants watching videos. This study represents an innovative investigation into quantifiable gaze patterns of adolescents with ASD during live social interaction vs. passive viewing of recorded social information.
Objectives:
To determine whether adolescents with and without ASD modulate their gaze patterns with social context.
Methods:
Adolescents (mean age 13:6, 17 ASD, 22 TD) watched four 60-second videos of TD adolescents talking about family, hobbies, vacation, and school. ASD diagnosis was confirmed via ADOS-2 and groups did not differ on age, IQ (KBIT, IQ>80), language ability (CELF), or sex. After several distractor tasks, participants sat across a table from a research assistant (RA), answered questions on the same four topics and also formulated their own questions to the RA (Winner et al. 2002). We used an SMI remote eyetracker to record the percent dwell time to the face during passive viewing (video) and the live interview.
Results:
Both groups gazed longer at the video face than the live face of the RA (p = .002), indicating that individuals in both groups similarly and appropriately reduced gaze duration (i.e., less staring) when facing a conversation partner compared to watching a video. Gaze patterns of TD adolescents for both tasks were significantly correlated (p = .01), suggesting internally consistent gaze strategies for social stimuli. Participants with ASD did not show this correlation (p = .23). We calculated between-task difference scores for gaze duration to ascertain each person’s modulation in gaze behavior as a function of task. Compared to the TD group, the ASD group showed less difference in looking time to the recorded video vs. the live interview (p = .06), maintaining similar gaze patterns for both conditions, while TD participants increased their gaze significantly more in the video task. Greater social communication impairment (SCQ) among ASD participants was correlated with decreased task-based modulation in gaze (p= .02; see Figure). The gaze patterns of TD participants were not correlated with SCQ scores.
Conclusions:
Both groups gazed longer at the video than the live person, but individuals with ASD modulated their gaze behavior less between the two tasks than TD peers. Greater social communication deficits in adolescents with ASD were correlated with increased dampening of this task-based gaze change, suggesting that reduced social insight may prevent individuals with ASD from fully modulating their gaze behavior for social stimuli. It is important to point out that the two groups differed most in gaze behavior to the video stimulus, not the live interaction, indicating that prior reports of significant eyetracking differences in ASD using video stimuli may not be describing the true social gaze behavior of children who have preserved language and cognitive abilities. More sophisticated and ecologically valid methods are required to quantify the foundations of aberrant social gaze during live interaction in this population.