A Gaze Contingent Exploration of Social Attention in Autism Spectrum Disorder

Background:

Attending to faces is extremely important for understanding social cues and it is thought that atypical social attention may contribute to social difficulties in Autism Spectrum Disorder (ASD). However, research findings are mixed as to whether faces have the same special status in attention for people with ASD as in typical development. Methodological differences across studies make it hard to reach a consensus about social attention in ASD and why it may appear typical or atypical in different studies. Additionally, the Social Motivation Theory of autism predicts that individuals with ASD will attend to non-social objects of particular interest to them, rather than faces, and previous research has found that objects of Circumscribed Interest (CI) may disrupt social attention when co-occurring with people in a scene.

Objectives:

The present study employed visual scenes containing people and objects of CI to which adults with ASD have previously been found to attend atypically. A novel gaze contingent window technique explored the allocation of attention across these scenes among adults with and without ASD. This technique provides a measure of top down allocation of attention by removing opportunities for automatic attentional capture in peripheral vision.

Methods:

In Experiment 1, a gaze contingent window revealed part of a scene containing people and CI objects within 4˚ visual angle around the point of fixation, and moved with the participants’ eye movements. In Experiment 2, participants saw an array of gift boxes, which, upon fixation, revealed a different image of either a face, an object of CI, or a neutral object. Frequency and duration of fixations within each area were recorded. Participants also later rated how much they liked each image and how interesting they found it to explore explicit interest in the stimuli.

Results:

Preliminary results (ASD: n=20 out of 30; Control: n=4 out of 30) found that all participants in Experiment 1 fixated for longer on faces within scenes compared to body or CI regions (ps < .04). In Experiment 2, participants with ASD had longer fixations to neutral objects compared to faces (p=.05). The frequency of fixations on each category of objects did not differ. In addition, participants with ASD rated faces as significantly less interesting and likeable than neutral or CI objects.

Conclusions:

Experiment 1 showed that when peripheral visual information is removed from a scene, social attention appears typical in adults with ASD, with faces holding attention longer than other objects once fixated. However, Experiment 2 found evidence for a reduction of social attention in participants with ASD, when objects were viewed without any meaningful context. This could be because social cues are more contextually relevant when trying to form an overview of a natural scene that is occluded (Experiment 1), compared to when there is no relation between hidden images (Experiment 2). This suggests that adults with ASD are able to preferentially allocate information to social cues when distracting information is removed, and it is beneficial to do so. Further time course analyses will help to elucidate these results.