The Role of Theory of Mind on Social Information Processing in Children With Autism Spectrum Disorders: A Mediation Analysis.

Mazza et al. (2017) asked why children with autism struggle with social cues. They tested if weak theory-of-mind skills sit in the middle of the problem.

Kids with autism and typical kids completed false-belief tasks and social-processing games. The team then ran a mediation model to see if ToM scores carried the link between diagnosis and social trouble.

The children with autism scored lower on both ToM and social-processing tasks. The stats showed ToM skill explained the social gap; when ToM was held constant, diagnosis no longer predicted social scores.

In plain words, mind-reading gaps drive the social-cognition gaps.

Chiu et al. (2023) extends the same idea over time: early ToM scores forecast better social play two years later in a large sample of autistic kids.

Begeer et al. (2015) seems to contradict the takeaway. They taught ToM lessons and saw gains on false-belief tests, but parents saw no wider social change. The gap is methodological: the RCT measured real-life social behavior, while Monica’s study used lab social-processing tasks that line up more closely with ToM content.

Frazier et al. (2023) offers a fix: they bundled ToM lessons with social-skills drills and quickly boosted both domains in three adolescents, showing the mediation path can be tapped when you train both pieces together.

You can stop treating social deficits as one giant target. Screen ToM first; if it’s low, add perspective-taking modules before or alongside social-skills training. Pair the two, as W et al. did, for faster change. And plan for anxiety supports—Chen et al. (2026) found better ToM can increase peer contact but also momentary social stress, so teach coping statements too.