A Psychometric Evaluation of the Diagnostic Analysis of Nonverbal Accuracy - Second Edition in Children with High Functioning Autism Spectrum Disorder
MetadataShow full item record
Children with high-functioning autism spectrum disorder (HFASD) are known to exhibit deficits in facial-emotion recognition (FER). However, research regarding FER in individuals with HFASD has yielded mixed results, potentially due to differences in task demands across various instruments. For this reason, it is important that assessment tools measuring this construct are reliable and well-validated. One popular measure of FER is the Diagnostic Analysis of Nonverbal Accuracy – Second Edition (DANVA-2; Nowicki & J. Carton, 1993). This computer-based task has been utilized in descriptive and clinical research studies in HFASD, but little work investigating its psychometric properties has been conducted by researchers other than the test developers. No independent studies examining the psychometric properties specifically within an HFASD sample were identified. Thus, the purpose of this study was to examine the psychometric characteristics of the DANVA-2 in this population. Specifically, the internal consistency, test-retest reliability, and construct validity of the DANVA-2 were assessed in a sample of 133 children with HFASD between the ages of 6 and 13 years. Test-retest reliability was assessed across immediate, 5-week, and 12-week retest intervals. Construct validity was evaluated by comparing it to concurrently administered tests of FER, measures of social skills and social competence, and measures of cognitive and verbal ability. Results indicated that, in an HFASD sample, the DANVA-2’s internal consistency reliability and test-retest reliability were adequate for research purposes, and its scores were highly convergent with other measures of facial emotion recognition. However, DANVA-2 scores were found to be unrelated to theoretically correlated constructs of social competence and autism symptomology. Study strengths and limitations along with implications for future research are discussed.