An advanced test of theory of mind: understanding of story characters' thoughts and feelings by able autistic, mentally handicapped, and normal children and adults.
Passing false-belief tasks does not mean real-life social understanding is intact.
01Research in Context
What this study did
Carr (1994) tested 89 people. Groups were able autistic, intellectually disabled, and neurotypical.
Each person read short stories about characters. They then explained what the characters thought and felt.
The stories were natural, like real life. This was harder than classic false-belief tasks.
What they found
Even autistic people who passed standard ToM tests scored lower on the stories.
They missed subtle thoughts and feelings that controls caught.
The gap stayed large for both children and adults with autism.
How this fits with other research
Smith et al. (2010) got the same result using see-know tasks. Both studies show moderate ToM gaps in autism.
Schuwerk et al. (2015) used eye-tracking and found adults with ASD still fail implicit false-belief tasks. Their deficits can improve fast with feedback, but they start behind.
Song et al. (2024) showed autistic kids also struggle with irony. Second-order ToM, not memory or attention, drives the problem. This lines up with G’s deeper-story deficit.
Richman et al. (2001) found high-functioning autistic kids miss pragmatic inferences in speech. Together, these papers say passing lab tasks does not mean real-world social thinking is fixed.
Why it matters
If a client passes Sally-Anne but bombs on social stories, keep teaching. Use naturalistic role-play, video modeling, and real-life inference drills. Track progress on everyday social questions, not just classic tests.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Add one short story-based inference probe to your session and score how well the client explains each character’s hidden feelings.
02At a glance
03Original abstract
Research has suggested that the core handicaps of autism result from a specific impairment in theory of mind (ToM). However, this account has been challenged by the finding that a minority of autistic subjects pass 1st- and even 2nd-order ToM tests while remaining socially handicapped. In the present study, able autistic subjects who failed ToM tasks, those who passed 1st-order, and those who passed 2nd-order tasks were tested with a battery of more naturalistic and complex stories. Autistic subjects were impaired at providing context-appropriate mental state explanations for the story characters' nonliteral utterances, compared to normal and mentally handicapped controls. Performance on the stories was closely related to performance on standard ToM tasks, but even those autistic subjects who passed all ToM tests showed impairments on the more naturalistic story materials relative to normal adult controls.
Journal of autism and developmental disorders, 1994 · doi:10.1007/BF02172093