Using the big data approach to clarify the structure of restricted and repetitive behaviors across the most commonly used autism spectrum disorder measures.

Background: Restricted and repetitive behaviors (RRB) in autism spectrum disorder (ASD) encompass several distinct domains. However, commonly used general ASD measures provide broad RRB scores rather than assessing separate RRB domains. The main objective of the current investigation was to conduct a psychometric evaluation of the ability of the Social Responsiveness Scale (SRS-2), the Social Communication Questionnaire (SCQ), the Autism Diagnostic Interview-Revised (ADI-R) and the Autism Diagnostic Observation Schedule (ADOS) to capture different RRB constructs.

Methods: Exploratory Structural Equation Modeling (ESEM) was conducted using individual item-level data from the SRS-2, SCQ, ADI-R and the ADOS. Data were obtained from five existing publicly available databases. For the SRS-2, the final sample consisted of N = 16,761 individuals (M = 9.43, SD = 3.73; 18.5% female); for the SCQ, of N = 15,840 (M = 7.99, SD = 4.06; 18.1% female); for the ADI-R, of N = 8985 (M = 8.86, SD = 4.68; 19.4% female); and for the ADOS, of N = 6314 (M = 12.29, SD = 6.79; 17.7% female).

Results: The three-factor structure provided the most optimal and interpretable fit to data for all measures (comparative fit index ≥ .983, Tucker Lewis index ≥ .966, root mean square error of approximation ≤ .028). Repetitive-motor behaviors, insistence on sameness and unusual or circumscribed interests factors emerged across all instruments. No acceptable fit was identified for the ADOS.

Limitations: The five datasets used here afforded a large as well as wide distribution of the RRB item scores. However, measures used for establishing convergent and divergent validity were only available for a portion of the sample.

Conclusions: Reported findings offer promise for capturing important RRB domains using general ASD measures and highlight the need for measurement development.