Verbal Intent and Gaze Cues on Action Prediction in ASD

A 2 (intent word: like, dislike) × 2 (object type: mentioned, unmentioned) × 3 (participant group: ASD, TD, ID) mixed experimental design was adopted. Children from the three age-matched groups (29 TD, 22 ASD, 24 ID) completed a video-watching task. In the video, a model was presented with two objects and uttered a sentence containing an intent word, a verb, and a noun, such as "I like to eat oranges." Based on the intent word, trials were divided into "like" and "dislike" conditions. Objects were classified as "mentioned" or "unmentioned" according to the verb and noun used. Participants were required to judge which object the model would pick while watching the video. Their choice responses and eye movements were recorded.

Inclusion criteria

• Children clinically diagnosed with Autism Spectrum Disorder (ASD). Typically developing (TD) children with no developmental or neurological disorders.

Children clinically diagnosed with Intellectual Disability (ID).

Common criteria for all groups: Normal or corrected-to-normal vision, no color vision deficiencies, normal hearing, and written informed consent provided by their legal guardians.

Exclusion criteria

• Failure to meet the study's group-specific screening cutoffs on the Social Responsiveness Scale (SRS) and Social Communication Questionnaire (SCQ).

Failure to meet the study's group-specific screening cutoffs on the Autism Diagnostic Observation Schedule (ADOS).

Known genetic syndromes, significant hearing loss, or uncorrectable visual impairment.