Open Collections

Abstract

Many autistic people experience challenges with social cognition (i.e., understanding others’ thoughts, feelings, intentions). Social anxiety (particularly evaluation-related fears), which frequently co-occurs with autism, has been linked to excessively negative mental state attributions and “reading too much” into others’ thoughts and feelings among allistic (i.e., non-autistic) samples. By contrast, some autistic individuals experience non-evaluative social fears which may be partially explained by difficulties recognizing others’ mental states. How these varied social-cognitive biases (over- vs. undermentalizing, negative interpretation bias) relate to different forms of social anxiety in autism warrants investigation. This dissertation examined (1) how social anxiety relates to specific social-cognitive errors (overmentalizing, undermentalizing, total errors) on a naturalistic measure of social cognition in autistic young adults, and whether these associations are moderated by evaluation concerns; (2) whether the relationship between evaluative social fears and mentalizing errors is partially explained by negative interpretation bias; and (3) whether conceptual measures (cognitive mentalizing, emotion recognition) and self-reported mentalizing skills suggest a similar pattern of findings as naturalistic measures. Sixty-four autistic young adults completed a multimethod battery of experimental and self-report measures assessing multiple facets of social cognition, social anxiety (social fear/avoidance, evaluation-related fears), autistic traits, and cognitive functioning. Findings revealed distinct and opposing social-cognitive profiles among autistic young adults with greater evaluative fears, who were generally more accurate on conceptual mentalizing measures and had higher self-reported mentalizing skills yet tended to over-interpret others’ mental states on a naturalistic measure, and those with greater autistic traits, who were more likely to make inaccurate mental state interpretations on a naturalistic task due to missing social cues and had lower self-reported mentalizing skills. Autistic traits and social fear/avoidance were highly interrelated, each independently associated with less accurate social-cognitive skills, potentially reflecting a distinct presentation of non-evaluative social anxiety amongst autistic individuals. Findings extend the current understanding of mechanisms underlying social anxiety in autism, indicate a need for tailored interventions to address these distinct presentations of social fears and social-cognitive differences amongst autistic adults, and encourage future research examining how social experiences and coping strategies may interact with social-cognitive differences to contribute to social anxiety in autism.