What did I say? Versus what did I think? Attributing false beliefs to self amongst children with and without autism.

The team asked kids with autism, kids with delays, and typical kids two kinds of questions.

First: "Where will the puppet think the toy is?" Second: "Where did YOU think the toy was a minute ago?"

The task removed extra words and pictures so only belief-tracking was tested.

Children with autism got the other-person question right more often than the self question.

The gap did not show up in the other groups.

This hints that spotting your own past false belief is extra hard in autism.

Colle et al. (2007) already showed that non-verbal false-belief tasks pick out autism from language impairment. Cramm et al. (2009) now add that the self-version of such tasks is even tougher.

Jellema et al. (2009), also from 2009, found autistic kids miss automatic social cues. Together these papers line up: both automatic and reflective social thinking can lag in the same diagnosis.

Bertschy et al. (2020) later showed that autistic adults categorize themselves less into groups. The self-belief struggle seen by M et al. may be an early sign of this broader self-other mapping issue.

When you test theory-of-mind, compare self and other questions. A child who passes standard false-belief probes may still fail to reflect on their own past thoughts. Adding a quick "What did you think first?" probe gives you a clearer picture and helps write self-monitoring goals that actually fit the learner.