Golnoosh Soroor¹, Alexandra McGowan¹, Amaya Aquino Aquino¹, Nicolas Davidenko¹; ¹University of California, Santa Cruz

Emotion recognition is crucial for functional social interaction. In some conditions, such as autism spectrum disorder (ASD), difficulties in emotion recognition have been reported and may be related to different gaze strategies (e.g., eye avoidance) and atypical perceptual processing, including a local processing bias. However, it is not clear whether variation in autistic traits across the general population predicts gaze and emotion recognition differences. In this study, ninety-five neurotypical participants completed an emotion recognition task evaluating carefully controlled face stimuli across the six basic emotion categories (anger, fear, disgust, happiness, sadness, and surprise). Before each block of trials, participants completed a Navon task designed to prime global processing, local processing, or intermediate processing (control). Eye tracking was used to measure gaze patterns across six basic emotions, including fixation distribution and dwell time in different facial regions. Participants then completed the Autism Quotient (AQ; Baron-Cohen et al., 2001) to assess autistic traits. Behavioral results showed that priming did not lead to differences in accuracy or reaction time. However, both global and local priming resulted in more broadly distributed gaze patterns, reflected by higher standard deviations of fixation locations, compared to the control condition. Across all participants, fixations were consistently more focused on the eyes than the mouth, regardless of AQ scores. However, participants with higher AQ scores (particularly in the attention-switching subscale) spent less time fixating on non-face areas. Overall, our findings suggest that the variability of AQ scores across the general population may not be sufficient to detect behavioral or gaze differences in emotion recognition previously shown in autism. Future research on emotional processing differences across the general population may need to examine emotion processing in more naturalistic contexts using dynamic faces.