Methods: The instrument is a 41-item Likert scale form parental questionnaire designed to screen ASD among children aged 6 years old-12 years old. It was filled out by the parents of 700 children (children with ASD = 47; typically developing children = 653) who were contacted through five special education schools and seven mainstream primary schools.
Results: Findings showed that Malay AQ-Child has an internal consistency of 0.82 as a whole scale for 41-items. Receiver operator characteristic analysis showed that the cut score of 63 for the translated, reduced version of the AQ-Child with 41 items had 99% area (95% confidence interval [CI]) under receiver operating characteristic (ROC) curve, and sensitivity and specificity of 93% and 99%, respectively. As for factor structure, principal component analysis (PCA) showed good loading values for most of the items in the instrument.
Conclusion: The translated version of the Malay AQ-Child for screening ASD can further facilitate the process of surveillance and early intervention for children who need it.