The Mandarin Chinese version of the childhood autism spectrum test (CAST): test-retest reliability.
Mandarin CAST scores stay stable for months, so you can trust them for autism screening in Chinese-speaking children.
01Research in Context
What this study did
Sun et al. (2013) asked a simple question: if a parent fills out the Mandarin CAST today, will they give the same answers two to four months later?
They gave the 37-item parent questionnaire to Chinese schoolchildren, then gave it again later. No teaching or therapy happened between the two times.
What they found
Scores stayed in the same place. The kappa value was 0.53–0.64 and rho was 0.73. In plain words, the Mandarin CAST is steady enough for research and screening.
How this fits with other research
Allison et al. (2007) ran a similar retest with the English CAST. Their kappa was a bit lower (0.41), but the pattern is the same: moderate stability, not perfect.
Sun et al. (2014) later showed the Mandarin CAST also has a clean two-factor structure and beats the older CABS screen. The 2013 reliability work laid the ground for those next steps.
Fitzpatrick et al. (2017) did the same job in Spanish. All three language versions now show the same message—CAST scores hold still over time.
Why it matters
If you screen Chinese-speaking kids, you can trust the Mandarin CAST at two-month intervals. A stable score means a high follow-up score is real, not noise. Use the same cut-off (15) and re-screen only after a few months unless clinical signs change.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Give the Mandarin CAST once, then re-screen no sooner than two months unless new red flags appear.
02At a glance
03Original abstract
This study aimed to investigate the test-retest reliability of a Mandarin Chinese version of the Childhood Autism Spectrum Test (CAST), in a Chinese population. Parents in a school based study on the prevalence of ASC in mainland China were asked to complete a second version of the CAST approximately 2-4 months after first completion. Test retest data were available from 70 children (questionnaires completed by the same parent). Using a cut-off score of 15, the test-retest reliability was good (kappa=0.64). The test-retest reliability in three categories (≤ 11, 12-14, ≥ 15) was moderate (weighted kappa=0.53). The correlation between the scores at CAST-1 and CAST-2 was good (Spearman rho=0.73). The Mandarin CAST demonstrated moderate to good test-retest reliability as a screening instrument for ASC in an assessment sample in mainland China.
Research in developmental disabilities, 2013 · doi:10.1016/j.ridd.2013.05.042