Anthropomorphic bias found in typically developing children is not found in children with autistic spectrum disorder.

Thierry et al. (2015) showed kids two side-by-side animations. One paired a human shape with human-like motion. The other mixed a human shape with robot motion.

Each child picked the clip they liked best. The team tested 20 autistic kids and 20 typical kids matched for age and smarts.

Typical kids chose the matched clip 80 % of the time. They liked human shape plus human motion.

Autistic kids picked at random. The match-up did not sway them. They showed no 'uncanny valley' bias.

Krüger et al. (2018) saw the same blank response. Their ASD group rated happy walking figures as less intense and felt less sure. Both labs find motion feelings land differently in autism.

Fink et al. (2014) looked at faces, not bodies. Once verbal skill was equal, autistic kids named emotions just as fast and as right as peers. The new body-motion task, however, still shows a gap. Face and body channels may not break together.

Capio et al. (2013) add a twist. In clinic tasks the ASD kids looked empathic, but parents still reported fewer real-life caring acts. The missing motion bias here may help explain that parent-clinic split.

If a child does not tune to natural body-motion cues, social stories that lean on those cues may fall flat. Try adding explicit labels or freeze-frame videos while you point out the feelings. Check parent report too—lab scores can look fine while daily life still feels off.