Cross-Cultural Revision and Psychometric Properties of the Chinese Version of the Autism Spectrum Rating Scale (2–5 Years)

Background: No sufficient biomarkers are available for early identification of autism in the general population. Currently, the diagnosis of ASD depends on behavioral assessments. A useful screening tool can help to detect early autistic symptoms and provide children an early opportunity for ASD-related interventions. This research aimed to assess cross-cultural adaptation and psychometric properties of the autism spectrum rating scale (ASRS) under the Chinese cultural environment. Methods: Participants were recruited from 17 kindergartens and 5 special education schools across five cities (Shanghai, Guangzhou, Changsha, Chengdu, and Harbin) in China. A total of 2,181 kindergarten children and 207 ASD cases participated in this study. Mplus 7.03 was utilized to conduct exploratory factor analysis, followed by adaptive modifications to construct the revised Chinese version of the ASRS (RC_ASRS). Results: The result showed that 62 items comprised a two-factor structure; Factor 1 (social communication, SC) included 21 items, and Factor 2 (unusual behavior, UB) included 41 items. Cronbach's alpha ranged from 0.87 to 0.91 within the RC_ASRS. The total score and the SC and UB scores were significantly higher in ASD cases than in kindergarten samples (Cohen's d ranged from 0.82 to 2.72). The total RC_ASRS score showed an area under the curve (AUC) of 0.95 (95% CI: 0.93–0.97). With a total score cut-off ≥ 60, the RC_ASRS is an excellent tool to identify ASD cases from Chinese kindergarten children (sensitivity = 88.6%, specificity = 84.5%). Conclusions: The RC_ASRS has excellent psychometric properties and is a reliable, useful tool for early ASD screening among Chinese children.