To examine whether the spatiotemporal organization of gaze strategies differs between children with autism spectrum disorder (ASD) and typically developing (TD) children during a staged joint attention interaction, and to identify at which phase differences emerge.
This cross-sectional observational study involved children with ASD and their typically developing (TD) peers freely viewing dynamic social interaction videos. The videos were segmented into four theory-driven phases: Introduction (scene establishment), Joint Attention (mapping another's gaze to an object preference), Intentional Comprehension (inferring the model's intention), and Emotional Feedback (evaluating outcomes that were either congruent or incongruent with the established preference). Participants' eye movements were recorded throughout using a remote eye-tracker.
Study Type
OBSERVATIONAL
Enrollment
66
School of Foreign Languages, University of Electronic Science and Technology of China
Chengdu, Sichuan, China
Percentage of Total Fixation Duration on Human Faces during Entire Video
Percentage of total fixation duration on human faces was calculated as (total fixation duration on all face areas of interest / total fixation duration on the entire video) × 100. Fixation was defined as gaze maintained within a 1° visual angle for a minimum of 100 ms. This measure was compared between the ASD and TD groups.
Time frame: During the full 27-second video presentation
Total Fixation Duration on Human Faces during Entire Video
Total fixation duration (in milliseconds) on human faces, calculated as the sum of all fixation durations that fell within face areas of interest (AOIs). This measure was compared between the ASD and TD groups.
Time frame: During the full 27-second video presentation
Percentage of Total Fixation Duration on Human Faces During Phase 4 (Emotional Feedback) by Condition
Percentage of fixation duration on faces was calculated for the emotional feedback phase separately for Preference-congruent and Preference-incongruent trials. The interaction between Group (ASD vs. TD) and Condition (Congruent vs. Incongruent) was examined. Higher percentages indicate greater allocation of visual attention to faces during emotional feedback.
Time frame: During seconds 22-27 of the video
Total Scanpath Length During Entire Video
Time frame: During the full 27-second video presentation
Number of Gaze Shifts Between Face and Object Areas of Interest
The number of saccadic gaze shifts that cross between pre-defined face and object areas of interest (AOIs) was counted. This was taken as an index of attentional switching between social and non-social information.
Time frame: During the full 27-second video presentation
Mean Distance of Fixation Points from the Center of the Face Area of Interest
This platform is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare professional.
For each fixation, the Euclidean distance (in degrees of visual angle) from its location to the centroid of the pre-defined face AOI was calculated. The mean of these distances across all fixations was then computed. Larger values indicate that gaze tended to fall farther from the faces, reflecting less attention to facial information.
Time frame: During the full 27-second video presentation