Role taking and social competence in autism and mental retardation.

Wehman et al. (1989) compared autistic youth to peers with the same mental age. They gave three role-taking tasks and several social-competence checklists.

The goal was to see if social problems come from general delays or from autism-specific gaps.

Autistic youth scored lower on every social-competence measure. They also scored lower on one of the three role-taking tasks.

The one role-taking task that did show a gap was the same task that lined up with social-competence scores.

Geurts et al. (2008) ran a close cousin study. They used story-telling instead of pictures and found the same thing: autistic kids could state the rules but still failed to shift perspectives.

Gillespie-Lynch et al. (2019) pooled 133 samples in a meta-analysis. The small effect they report now wraps around the 1989 data point, showing the field has known about this gap for decades.

Callenmark et al. (2014) looked deeper and found a twist: autistic teens pass explicit perspective tests yet still fail spontaneous ones. This explains why the 1989 paper saw a deficit on only one task—the tasks that felt natural, not quiz-like.

If you run social-skills groups, do not assume a child who can parrot social rules truly takes another’s view. Add quick, real-time perspective checks: pause a video and ask, "What does she think just happened?" Use natural tasks, not worksheets. Target the skill that actually links to peer problems.