Unusual Gaze Direction Detection in Autism Spectrum Disorders.
Autistic teens adapt to gaze direction just like peers, so basic gaze sensing is intact.
01Research in Context
What this study did
Feng et al. (2025) asked if autistic teens process eye-gaze direction differently.
They used a gaze-adaptation task while recording brain waves.
Both autistic and neurotypical teens stared at faces whose eyes looked left or right.
What they found
The brain waves showed no group difference.
Autistic teens adapted to gaze direction the same way their peers did.
This result challenges the idea that autism involves faulty gaze prediction.
How this fits with other research
Fletcher-Watson et al. (2008) and McLennan et al. (2008) also found intact gaze cueing in autistic adults, so the null result extends to teens.
Uono et al. (2021) saw equal accuracy for detecting self-directed gaze, matching the new null finding.
Akechi et al. (2014) seems to disagree: they reported that autistic teens lack unconscious eye-contact bias.
The clash is only skin-deep. Tianqi tested conscious adaptation; Hironori tested unconscious detection. Different brain levels, different answers.
Why it matters
You can stop assuming teens with autism miss basic gaze cues.
Teach social use of eye contact, not detection of it.
Add clear rules and scripts instead of drilling gaze following.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Pair eye-contact goals with explicit rules ("look when you hear your name") instead of cueing drills.
02At a glance
03Original abstract
Predictive processing accounts of autism posit that autistic individuals' perception is less biased by expectations than nonautistic individuals', perhaps through stronger precision-weighting of prediction errors. Since precision-weighting is fundamental to all information processing, under this theory, the differences between autistic and nonautistic individuals should be domain-general and observable in both behavior and brain responses. This study used EEG, behavioral responses, and eye-tracking co-registration during gaze-direction adaptation, to investigate whether increased precision-weighting of prediction errors is evident through smaller adaptation after-effects in autistic adolescents compared with nonautistic peers. Multilevel modeling showed that autistic and nonautistic adolescents' responses were consistent with behavioral adaptation, with Bayesian statistics providing extremely strong evidence for the absence of a group difference. Cluster-based permutation testing of ERP responses did not show the expected adaptation after-effect but did show habituation to repeated stimulus presentation, and no group difference was detected, a result not consistent with the theoretical account. Combined with the few other available studies, the current findings raise challenges for the theory, suggesting no fundamental difference in precision-weighting of prediction errors in autism.
Journal of autism and developmental disorders, 2025 · doi:10.1002/aur.3118