Psychometric Properties of Persian Version of Checklist for Autism in Toddlers (CHAT)

Abstract

Background: The increasing rate of autism spectrum disorder highlights the need for early screening with the aim to provide early intervention. This study aimed to provide Persian version of Checklist for Autism in Toddlers (CHAT) and assess its psychometric properties such as validity and reliability. Methods: This was a cross-sectional study with psychometric design on 318 children (97 with and 221 without autism) selected from training specialized centers for autism children and psychology clinics in Isfahan City, Iran during the years 2016-2017. The subjects were chosen according to inclusion criteria using convenience sampling method. The using scale was Persian version of CHAT. To determine its reliability, Cronbach's alpha method was used; and to test the construct, concurrent, and discriminate validity, factor analysis and its correlation with Modified Checklist for Autism in Toddlers (M-CHAT), Pervasive Developmental Disorders Screening Test-Second Edition (PDDST-II), and Psychiatric diagnosis were used. The data were analyzed using correlation, factor analysis, discriminant analysis, logistic regression, and linear regression. Findings: The Cronbach's alpha was 0.88, and the correlation of items with total scores ranged from 0.35 to 0.71. Concurrent validity, by correlations with M-CHAT, PDDST-II, and psychiatric diagnosis, was 0.6, 0.79, and 0.57, respectively. Factor analysis revealed 2-factor structure (social-emotional and interactive behavior), and discriminant analysis in the cutoff point 2.5, delineated the sensitivity of 67% and the specificity of 91%. Conclusion: The Persian version of CHAT, as a preliminary screening tool for 18-24-month-old children with autism, is capable of proper psychometric values. However, newborns who are at risk should be reassessed again several months after the age of 24 month.

Keywords

• Reliability and validity
• Checklist
• Autism
• Sensitivity
• Specificity