Video Continuation Prediction in Autistic and Non-Autistic Adults

In this study, we seek to examine the ability of autistic adults recruited through SPARK and non-autistic adults recruited through Prolific to predict which of two video sequences is more likely to follow a preceding one. On each of 15 trials, participants view a base video followed by two AI-generated continuation videos and use a continuous slider to indicate how likely the left vs. right continuation video is to follow the base video. One continuation video per trial is intended to be more plausible given the base video. The primary data are the slider responses. We hypothesize that non-autistic participants will show higher accuracy, more extreme responses in the direction of the intended continuation video, and greater between-subject consistency than autistic participants.