Psychometric evaluation of the Autism Symptom Dimensions Questionnaire.

AIM

To replicate and extend previous psychometric findings for the Autism Symptom Dimensions Questionnaire (ASDQ).

METHOD

Using a cross-sectional design in two samples, including a total of 3366 children and adolescents (aged 2-17 years; 1399 with autism spectrum disorder) and a small case series, factor structure, measurement invariance, reliability, construct validity, screening and diagnostic efficiency, and detection of reliable change were examined for the ASDQ.

RESULTS

Strong psychometric properties were observed, including replicable factor structure, strong measurement invariance, adequate-to-excellent scale and conditional reliability, strong convergent and discriminant validity, and good screening efficiency. Importantly, two crucial measurement aspects required for a good treatment outcome measure were found, that is, good test-retest stability and the potential to detect reliable change. As expected for an informant-reported questionnaire, diagnostic efficiency was weak.

INTERPRETATION

The ASDQ showed evidence of favorable and well-replicated psychometric properties in two samples. Good screening but weak diagnostic efficiency, coupled with good test-retest stability, suggest that optimal clinical use is in screening and in detailed clinical characterization beyond what is offered by current informant-reported measures, with the potential for monitoring response to intervention. Despite relative brevity, the ASDQ provides good coverage of broad and specific aspects of the autism behavioral phenotype, which is consistent with the criteria of the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition.