Judging Intentionality of Ambiguous Action in an Adult ASD Sample

Background: Discerning intentional from unintentional actions is a key aspect of social cognition. Mental state attribution literature has consistently shown that people with autism spectrum conditions tend to be less accurate in discerning an agent’s intention when there is clearly a right answer. However, little is known about how autistic people attribute intentions when making judgements about ambiguous actions. Understanding ‘intention attribution style’ in this population will help us gain an insight into everyday mental state processing whereby intentions may be frequently ambiguous.

Objectives: The aim of this study was to determine whether individuals on the autism spectrum differ in their interpretation of ambiguous actions compared to neurotypical controls.

Methods: Participants included 20 adults with a diagnosis of an autism spectrum condition (7 females) and 20 neurotypical controls (11 females). Verbal and perceptual reasoning ability was measured using the Wechsler Abbreviated Scale of Intelligence II. All participants were asked to complete a modified version of Rosset's (2008) Ambiguous Sentences paradigm which measures the proportion of intentional versus unintentional judgements of ambiguous actions.

Results: There were no significant differences in verbal ability or perceptual reasoning ability between groups. The Strange Stories Film task (SSFt; Murray et al., 2017) which measures advanced theory of mind (ToM) revealed a significant difference in ToM abilities between the autistic group and controls. One statistically significant outlier in the ASD group was excluded from analysis of the Ambiguous Sentence paradigm scores. Participants with autism showed a higher intentionality endorsement percentage for ambiguous sentences (M=18.42.91, SD=13.18) than controls (M=13.18, SD=8.59; t(37)=2.04, p=.048). ToM scores did not correlate with intentionality endorsement percentage scores in either group therefore group differences could not be explained by ToM ability.

Conclusions: Our results suggest that high functioning adults with an autism spectrum condition are more likely to perceive ambiguous behavior to be intentional rather than accidental compared to neurotypical controls. This is a noteworthy result in a small sample and suggests group differences in intention attribution style. Over-attributing intentions when interpreting ambiguous behaviour may be a contributing factor to the social difficulties experienced by autistic individuals. As ToM deficits could not explain the increased tendency to perceive ambiguous but prototypically accidental behavior to be intentional, a follow-up study investigating the relation between executive functioning skills and intention attribution style in autism is planned.