Convergent validity of the Autism Spectrum Disorder-Diagnostic for Children (ASD-DC) and Childhood Autism Rating Scales (CARS)

Previous studies analyzed the reliability as well as sensitivity and specificity of the Autism Spectrum Disorder-Diagnostic for Children (ASD-DC). This study further examines the psychometric properties of the ASD-DC by assessing whether the ASD-DC has convergent validity against a psychometrically sound observational instrument for Autistic Disorder. Thirty-seven children and adolescents, ages 3–14 years old were studied. Spearman's Rho was utilized to assess the convergent validity between the ASD-DC and the Childhood Autism Rating Scale (CARS). The ASD-DC and CARS total scores were compared. Additionally, the ASD-DC nonverbal communication, verbal communication, social relationships, and insistence on sameness/restricted interest factors were compared to the CARS items of nonverbal communication, verbal communication, relating to people, and adaptation to change. All correlation coefficients were statistically significant and ranged from .37 to .68, indicating that the ASD-DC demonstrated convergent validity against the CARS. Percent of correct classification was also computed for the ASD-DC and CARS as compared to diagnosis assigned for research purposes according to a checklist derived from the ICD-10 and DSM-IV criteria for autism spectrum disorders. On the ASD-DC, 76.5% correctly met cutoff for ASD and 95.0% were correctly classified as not meeting cutoff for an ASD. Of those meeting the cutoff on the CARS, 58.8% were correctly classified as having an ASD and 85.0% were correctly classified as not having ASD.