Preference for Nonsocial Realistic Movement in Children with ASD

Background: From infancy children orient to social information as supported by evidence of preference for and orienting to both faces and biological motion. In ASD, patterns of visual attention for socially based stimuli often differ from typically developing peers. In particular there is evidence for a lack of orienting to biological motion (Klin et al., 2009) and a preference for non-biological motion in the form of geometric patterns (Pierce, et al., 2011).

Objectives: To examine whether diminished orienting to motion is specific to socially based motion or a general diminished orienting to realistic motion.

Methods: Participants were 19 children (Female = 4) with ASD ranging in age from two to five years (M = 4.24, SD = 1.12). Stimuli consisted of a total of 20 5-second movie clips of social or object based motion with no sound. Social movies were male and female children in the age range of the study sample engaging in a variety of movements like jumping, dancing, and waving their arms. Object movies were a range of objects moving independently like a remote control bus driving, a record spinning, and marbles going down a marble tower. Control stimuli were created by blurring the social and object movies, retaining color, brightness and amount of movement, but were not obviously movement from everyday life. Participants viewed each clip simultaneously (side by side) with the blurred version of itself on a 22”” computer screen. Children’s gaze shifts were monitored with an SMI REDn Remote Eye-Tracking system. Looking time was measured as visual fixations, defined as ≥100 ms of continuous gaze to a 100 pixel area, and saccades within the visual area of each video presentation. Preference for the blurred video was calculated as looking time to the blurred video over looking time to both the blurred and real movement videos. Preference was averaged across 10 trials within each condition. Trials with less than 200 milliseconds of looking to either video were excluded.

Results: One-sample t-tests with a test value of .5 revealed that in the object condition, participants demonstrated a preference for the object movement versus the blurred video (M = 0.38, SD = 0.11, t(18) = -4.79, p < .0001). In the social condition participants’ preference was not significantly different from chance (M = .52, SD = 0.15, t (18) = 0.71, p = .49).

Conclusions: This paradigm demonstrated the presence of a preference for realistic motion when that motion is non-biological. These results suggest that orienting to biological motion may be a domain specific deficit. Next steps are to examine relationships between preference and measures of ASD symptoms, as well as to compare performance on this task to children with developmental delays and typically developing children.