Fixation Patterns to Faces in Autism: Investigating the Influence of Task Requirements

Background: Although reduced eye contact in autism is a clinically and anecdotally well-established phenomenon, experimental studies aimed at quantifying this phenomenon have yielded mixed results. In part, this may be explained by differences in the type of face stimuli used [for example, static vs. dynamic faces (Speer et al., Autism, 2007) and intact vs. degraded faces (Spezio et al., JADD, 2007)]. However, it remains unclear as to whether inconsistent findings might also be explained by differences in task demands rather than just the nature of the stimuli.

Objectives: Here we provide a detailed characterization of the role of task demands on scan patterns to faces in autism, as well as comparisons with control participants. By using identical face stimuli across a series of different tasks, and running each task on each participant, we determine the contributions of, and interactions between, task requirement and participant on scan patterns to faces.

Methods: While undergoing eye-tracking, subjects viewed smiling direct-gaze male and female face images while performing the following tasks, presented in a random order: (a) passive viewing, (b) gender discrimination, (c) 2-back gender discrimination, and (d) friendliness judgment. Images were presented to the subject in a gaze-contingent manner wherein subjects had to first fixate on a cross on one side of the screen before the image would appear on the opposite side of the screen. Analyses of interest include location and latency of first fixations, number of fixations, and the amount of time fixating various features of the face (e.g., eyes, mouth).

Results: Preliminary analyses of 5 adult participants with autism and 6 neurotypical control participants confirm that individuals with autism spend less time fixating the eye region of faces, relative to controls. Furthermore, across the various tasks, fixation patterns are remarkably consistent within an individual participant. However, there are differences across participants, with some participants consistently making more or less eye contact than other participants.

Conclusions: Thus far, results suggest that task requirements have little influence on how individuals with autism fixate faces. Rather, differences in scan patterns are seen between individuals. If these findings were confirmed in a larger group, this would suggest that the inconsistencies in the literature regarding reduced eye contact in autism are likely driven in large part by differences in the study participants, and possibly also differences in the type of stimuli used, rather than differences in task requirements. Data collection is ongoing, and the results from a larger group of participants will be reported.