Abstract
Our patterns of eye movements are influenced by task demands (Chen & Zelinsky, 2006). Malcolm et al. (2008) reported preferential scanning of the upper and lower face for information about identity and emotion, respectively. However, their data. were based on scanning of faces displaying only two emotions: happiness and disgust. Both of these emotions are predominantly expressed using the lower half of the face. We sought to determine whether Malcolm et al.’s conclusions would extend to the perception of more emotions. Additionally, we investigated whether adults with autism spectrum disorder (ASD), a clinical population with known face processing difficulties, showed similar eye movement preferences to those of matched controls. One set of stimuli was used comprising faces that expressed anger, disgust, fear, happiness, neutrality, sadness, and surprise while young adults made identity and emotion judgments. Eye-tracking, reaction time, and accuracy data were collected. Participants made more upper- than lower-face fixations for both the emotion and identity tasks. Like Malcolm et al (2008), the proportion of upper- to lower-face fixations was greater for the identity than the emotion task for all emotions. Relative reliance on the upper and lower regions of the face varied with emotion, regardless of the task. Participants with ASD scanned more differently during the emotion than the identity task. They fixated less often overall, yet still showed a general trend to fixate the upper more than the lower face. However, this pattern varied to a greater extent with face emotion than in typically developing participants, with some emotions revealing no preference for the upper or lower face. Results of this experiment confirm the presence of top-down, task-driven effects on the visual scanning of faces, effects that depend on facial emotion and that are altered in individuals with ASD.
Meeting abstract presented at VSS 2013