Behavior
Time Series Analysis to Decode Children’s Response to Stressful Social Stimuli Shriya Reddy* Shriya Reddy Shreeya Patel Devanshi Patel Gesulla Cavanaugh
Children with Autism Spectrum Disorder (ASD) frequently face challenges related to attention deficits, particularly when exposed to stressful situations that hinder their ability to concentrate and disrupt their attention spans. While numerous studies delve into the social responses of children with ASD, a notable gap exists in eye-tracking models designed to anticipate physiological reactions to stress within this demographic. This comprehensive two-part study adopts a time series methodology, analyzing facial expressions, eye movement, and pupil diameter data from children with ASD (n=17) in comparison to their neurotypical counterparts (n=13). Data in the initial stages were gathered by using the Tobii Pro Nano, semi-structured surveys, a computer-embedded camera, and the M-CHAT tool. For each stimulus from the generated videos, Areas of Interest (AOIs) were created. For each of these AOIs, pupil diameter, number of saccades, and fixation were collected, recorded, and calculated for all events of interest. To analyze this data, Tobii Pro lab and IBM SPSS V 27.1 were utilized. From all the recordings, the results suggest that children with ASD had greater unstable eye movement. The time series data of children with ASD revealed greater increases in pupil diameter at the beginning of each stimulus along with considerable decreases at the onset of stressful stimuli. No significant differences by group (p = 0.78) were noted for fixation duration on objects with no social connotation (alpha 0.05, 95% CI). These findings suggest that eye-tracking technology can be utilized to detect acute stress, attention, and disengagement in children with ASD, offering potential support for future therapy planning tailored to their needs.