Examining the Causal Mechanisms of Positively-Biased Self-Perceptions in Adolescents with Autism Spectrum Disorders

Background: Research has demonstrated that, despite problems in multiple domains, children with Autism Spectrum Disorders (ASD) show a lack of awareness of their difficulties. Previous research on the self-perceptions of children with ASD has focused primarily on examining perceptions of their autism-related traits by utilizing discrepancy analyses on broad, abstract measures. This method involves difference scores that are calculated by subtracting a criterion score (e.g., parent report) from the child’s self-report of competence. The present study aimed to overcome these potential biases by examining the self-perceptions of adolescents with ASD using concrete, familiar, and objective criteria to assess their understanding of academic capabilities.

Objectives: To extend research on self-perceptions held by adolescents with ASD. By investigating the relationships between IQ and self-perceptions and executive functioning and self-perceptions, the study hoped to provide a preliminary examination of the underlying causal mechanisms related to biased self-perceptions.

Methods: Forty-one participants, 19 with ASD and 22 typically-developing (TD) adolescents (age range =12-18 years), participated. ASD and TD groups were matched on mental age. Self-report questions were utilized after participants completed both verbal and mathematic tasks. Participants were asked how well they thought they did prior to completing the tasks (pre-task prediction). After they completed each task they were asked how well they thought they did (immediate post-performance) and how well they thought they would do in the future (hypothetical future post-performance) Difference scores between actual performance and predicted performance were analyzed. The Delis-Kaplan Executive Function System (D-KEFS) was used as a measure of executive functioning and the Wechsler Abbreviated Scale of Intelligence (WASI) was used as a measure of IQ.

Results: A one-way MANOVA was conducted to determine the effect of group diagnosis on the three dependent variables: pre-task prediction, immediate post-performance, and hypothetical future post-performance. Significant differences were found between groups on the dependent measures, Wilks’s Λ = .51, F(3,37) = 11.70, p < .001, η²= .49. Pairwise comparisons indicated that the ASD group had higher difference scores in pre-task prediction, immediate post-performance, and hypothetical future post-performance questions. Correlation coefficients among difference scores suggest that participants with higher IQ have a more accurate pre-task prediction. Correlation coefficients among difference scores and the D-KEFS suggest that participants with greater abilities in concept formation tend to have a more accurate pre-task prediction while participants with greater abilities in inhibition and cognitive flexibility tend to have more accurate immediate and hypothetical future post-performance self-perceptions.

Conclusions: Examining self-perceptions in ASD, using pre/post task self-competence questions, furthers our understanding of the causal mechanisms underlying this phenomenon. It is essential to understand more about the self-perceptions held by individuals with ASD, as a lack of awareness of poor competence may make it difficult for individuals to adjust their behaviour in accordance with feedback This lack of awareness may also limit their ability to incorporate feedback into an understanding of their functioning, thus leading to greater impairment over time. This study may help future research to develop strategies to deal with the potential negative implications associated with overly positive biases.