Abstract
Background. People generally look at the eyes when viewing faces, but people with autism tend to avoid the eyes. Measurements of these behavioral differences can be problematic, however, because people with high-functioning autism or Asperger’s syndrome often adopt top-down behavioral strategies that mask internal tendencies. We attempt to unmask these differences in conflict situations, exploiting Laidlaw, et al.’s (VSS’11) finding that unscreened participants spontaneously look at the eyes even when their task is to avoid doing so. Method. Unscreened student participants filled out the Baron-Cohen Autism Quotient (AQ) questionnaire and viewed a series of face images (FaceGen) while their gaze was monitored (Eyelink2). There were three blocks of 48 trials, each with different instructions: don’t look at the eyes ("avoid eyes"), don’t look at the mouth ("avoid mouth"), or "free view". Face images had three possible expressions: neutral, happy, or disgusted. Results. Comparing subjects with high AQ (>32; N=5) vs. low AQ (<22; N=10) for the proportion of time spent looking at the eyes and the mouth, no significant differences were found in the "free view" or the "avoid eyes" conditions. However, in the "avoid mouth" condition, the high AQ group looked at the eyes significantly less than the low AQ group did (t-test, p<0.03). Regression analysis (N=27) showed the same: time spent looking at the eyes was negatively correlated with AQ scores only in the "avoid mouth" condition (p<0.04). Disgusted faces tended to elicit a bigger group difference (ANOVA, p=0.06). Discussion. Excessive top-down inhibition in the high AQ group can be a factor in the "avoid eyes" and the "free view" conditions, but not in the "avoid mouth." Whereas the "avoid eyes" may measure strength of such task-relevant control, the "avoid mouth" gets around it to isolate the purely spontaneous tendency to avoid the eyes.
Meeting abstract presented at VSS 2012