Poster #215 - Visit Count Gaze Patterns to Emotional Faces Coincide with Individual Differences in Autistic Symptomatology

• In Event: 1-061 - Poster Session 02
  In Poster Session: PS 02 Section - Social, Emotional, Personality

Integrative Statement

Eye tracking provides a unique methodology for understanding emotional facial recognition in individuals with autism; however, a coherent, consistent exploration of all areas of the faces (upper, middle and lower; see Ekman & Friesen, 1976) has not been considered. Previous research has revealed individuals with autism fixate less on core features (only eyes and mouth) than non-core features (Pelphrey et al., 2002). Eye tracking has been used to analyze fixations, a measure of initial interest and emotion perception, but the use of an alternative eye-tracking variable, visit count, has not been investigated. Visit count has the potential to be more informative for emotion processing, rather than emotion perception, as it is indicative of increased interest in a particular AOI. Visit count provides a measure of gaze and dwell time to an AOI, suggesting greater visit count indicates greater interest, the focus of the current study. Thus the current study increases our understanding by exploring an under-represented dependent variable, visit count, and specific regions of the face important in emotion recognition.
Eighty facial expression images of five emotions (happy, sad, anger, fear, and neutral), equally distributed across sex, were taken from CAFE child facial set (LoBue & Thrasher, 2014) and NimStim adult facial set (Tottenham et al., 2009). Four areas of interest (AOI) were identified based on Ekman & Friesen’s (1976) emotion recognition literature: lower (nose and mouth), middle (eyes and cheekbones), upper (eyebrows and forehead), and non-core (external features relevant for emotion recognition e.g. hair, ears, chin). Faces were presented for four seconds while eye tracking technology measured visit (dwell time) patterns. Participants were asked to make emotion judgments for each face by rating them on two scales: arousal and pleasantness. Participants were asked to complete the Autism Spectrum Quotient questionnaire (AQ-50, Baron-Cohen, 2001), which was used to group participants into quartiles based on their autism quotient (AQ) scores; higher scores are indicative of greater autistic symptomatology.
A 2(age) x 5(emotion) x 4 (AOI) x 4 (AQ) repeated measures ANOVA was conducted on visit count. A three-way interaction among emotion, area of interest and AQ group approached significance, F (36, 852) = 1.405, p = 0.059, see graphs. In general, a clear distinction in visit frequency emerged between the lower and middle AOIs compared to the non-core and upper AOIs within the lowest and highest AQ scoring groups. Between AQ groups, the lower and middle AOIs appear to be of equal importance, which is contrary to findings from previous literature. Gaze pattern differences among the AQ groups emerged for anger and fear. Specifically, for fear, the lower had greater visit count in lower scoring AQ groups, but the middle had greater visit count in higher scoring AQ groups. For anger, the lower had greater visit count in lower scoring AQ groups, however, there was no distinction between the lower and middle in the higher scoring AQ groups. The use of visit count analyses may reveal patterns between typically developed individuals and individuals with ASD not revealed when analyzing fixations.

Authors

• Ayesha Ferozpuri, Christopher Newport University
  Presenting Author

• Laurie Sullivan Hunter, Christopher Newport University
  Non-Presenting Author

• Samantha Leigh, Christopher Newport University
  Non-Presenting Author

• Isabelle Clough, Christopher Newport University
  Non-Presenting Author

• Leslie Rollins, Christopher Newport University
  Non-Presenting Author