The psychometric properties of the Quantitative-Checklist for Autism in Toddlers (Q-CHAT) as a measure of autistic traits in a community sample of Singaporean infants and toddlers

I. Magiati, D. A. Goh, S. J. Lim, D. Z. Q. Gan, J. C. L. Leong, C. Allison, S. Baron-Cohen, A. Rifkin-Graboi, B F P. Broekman, S-M. Saw, Y-S. Chong, K. Kwek, P. D. Gluckman, S. B. Lim, M. J. Meaney, on behalf of the GUSTO working group

There is growing research evidence that subclinical autistic traits are elevated in relatives of individuals with autism spectrum disorder (ASD), continuously distributed in the general population and likely to share common etiology with ASD. A number of measures have been developed to assess autistic traits quantitatively in unselected samples. So far, the Quantitative-Checklist for Autism in Toddlers (Q-CHAT) is one of very few measures developed for use with toddlers as young as 18 months, but little is known about its measurement properties and factor structure. The present study examined internal consistency, factor structure, test-retest stability, and convergent validity of the Q-CHAT in a sample of toddlers in Singapore whose caregivers completed the Q-CHAT at 18 (n = 368) and 24 months (n = 396). Three factors were derived accounting for 38.1 % of the variance: social/communication traits, non-social/behavioral traits, and a speech/language factor. Internal consistency was suboptimal for the total and speech/language scores, but acceptable for the social/communication and non-social/behavioral factor scores. Scores were generally stable between 18 and 24 months. Convergent validity was found with the Pervasive Developmental Disorders subscale of the Child Behavior Checklist (CBCL) completed by caregivers when their children were 24 months. Q-CHAT total scores in this sample were higher than those reported in other unselected samples from the UK. The Q-CHAT was found to have a three-factor structure, acceptable internal consistency for its two main factor scores (social/communication and non-social/behavioral), normally distributed scores in an unselected sample, and similar structure and measurement properties as those reported in other published studies. Findings are discussed in relation to existing literature and future directions for the validation of the Q-CHAT.