Quantitative Checklist for Autism in Toddlers (Q-CHAT): A psychometric study with Serbian Toddlers

Abstract

BackgroundThe present study evaluated the underlying structure of the Quantitative Checklist for Autism in Toddlers (Q-CHAT) in a Serbian sample and its measurement properties for community-based screening. MethodData were collected from parents of 220 toddlers aged 15–36 months. The parents completed together the Q-CHAT and the Modified Checklist for Autism in Toddlers - Revised (M-CHAT-R). Confirmatory factor analysis (CFA), internal consistency, convergent and known-groups validity were analyzed, with sensitivity, specificity, and predictive values calculated. ResultsFrom CFA, 17 out of 25 items emerged to represent three underlying factors, namely social communication/interaction, restrictive/repetitive behavior, and language impairment. The Q-CHAT-10 was found to have one-dimensional structure. Cronbach’s α was ≥.81 for Q-CHAT scores. All scores for toddlers with typical development were significantly lower compared to those with ASD. The cut-off points that best-balanced sensitivity/specificity for the total scores, were 37 for the Q-CHAT with all 25 items (96.2/81.9 %), 26 for the Q-CHAT with 17 items (96.2/93.3 %), and 3 (100/93.3 %) for the Q-CHAT-10, with the positive/negative predictive values of .39/.99, .57/1.00, and .67/1.00, respectively. ConclusionsThe Serbian Q-CHAT versions with 10 and 17 items showed sound internal consistency, convergent and know-groups validity, and high levels of sensitivity and specificity for community-based screening. A multidimensional structure with social communication/interaction, restrictive/repetitive behaviors, and language impairment aspects was confirmed. This implies that autistic traits in toddlers could be measured dimensionally with the Q-CHAT, but its structure should be verified in additional studies.