Reduced Mimicry to Virtual Reality Avatars in Autism Spectrum Disorder.

The study used a 2D screen-based VR environment, and participants rated the avatars as not very real, so the social manipulation may have failed because the avatars felt unrealistic. The avatars’ hand movements were based on human motion capture, but their head movements and smiles were animated, which may have reduced realism. The task required participants to remember a sequence of three targets, and the error rate was higher than in simpler reaction-time tasks, so task demands may have masked social effects. The sample was 25 neurotypical and 26 autistic adults, with high average IQ scores, so it may not apply to children or people with more severe communication needs. The group difference in mimicry was marginal, so the study does not clearly establish reduced mimicry in autism. The paper did not test a therapy, patient outcome, or VR training programme, so it cannot tell clinicians whether VR improves social communication. Some autistic participants did not meet the overall ADOS classification for autism or autism spectrum, although they met a subscale cut-off.

Funding was provided by the European Research Council. The supplied text does not report other conflicts of interest or author disclosures.

Do not conclude that socially engaged VR avatars increase mimicry or that VR social-skills training is effective. The avatar’s engagement did not change copying, the co-presence ratings were low, and no clinical outcome was tested.