Convergence of Self-Report and Informant Measures of Executive Function for Adults with ASD

Background: The number of adults with Autism Spectrum Disorder (ASD) is rapidly increasing, and with it, the need for appropriate assessment instruments to better understand this population. Self-report measures are widely used for both research and clinical assessment of adults with ASD, though there is little research examining the convergence of self- and informant-report in this population.

Objectives: This research capitalized on a unique dataset of adults diagnosed with ASD as children by the TEACCH Autism Program from 1970-1999. The present study examined agreement between Caregiver- and Self-Report on the Behavior Rating Inventory of Executive Functioning–Adult Version (BRIEF-A; Roth & Gioia, 2005).

Methods: Caregiver surveys were completed for 274 adults with ASD in middle adulthood. Of this sample, 25 adults with ASD (age range: 27-57, M = 35.1) and their caregivers completed the Self- and Informant-Report versions of the BRIEF-A during in-person follow-up testing. The self-report form of the BRIEF-A was only administered to individuals with a full-scale IQ (FSIQ) of 70 or higher (FSIQ range: 73-120, M = 92.24).

Results: Caregiver and Self-Report t-scores on the BRIEF-A Global Executive Function Composite (GEC) were significantly correlated (r = .73, p < .01). Scores on the Behavioral Regulation and Metacognition indices also demonstrated significant, positive correlation (r = .75, p < .01; r = .64, p < .01, respectively). There was not a significant difference between Caregiver and Self-Report t-scores for the GEC (Caregiver M=56.28; Self-Report M=54.92). However, for two subscales, Self-Monitoring and Organization of Materials, Self–Report t-scores were significantly lower than Caregiver t-scores (t = 2.00, p < .05; t = -2.49, p < .05, respectively). Discrepancy scores were calculated by subtracting Self-Report t-scores from Caregiver t-scores on the GEC and on each subscale. Discrepancy size was significantly negatively correlated with the FSIQ of the adults with ASD. Larger discrepancies between Caregiver and Self-Report were associated with lower FSIQ scores (r = -.58, p < .01). Discrepancies were largest for individuals with FSIQs below 85, with the adults with ASD reporting significantly fewer executive function impairments than their caregivers reported about them.

Conclusions: The BRIEF-A demonstrated acceptable levels of inter-rater agreement when used to measure executive functioning for adults with ASD and did not show overall differences in t-scores for Caregiver and Self-Report version. However, cognitive ability was significantly related to the size of difference between Caregiver and Self-Report. Because individuals with lower FSIQs may have more difficulty accurately reflecting on their own impairments, under-reporting executive function issues may result in disqualification from services and, more broadly, an inaccurate understanding of executive function abilities in this population. For individuals with lower cognitive abilities, gathering information from multiple sources is important for a more complete assessment of functioning than gathering information from an adult or caregiver alone.