The latent structure of the Delis-Kaplan system for autism.

Abstract

A core feature of autism is deficits in executive functioning (EF), including difficulty with planning, cognitive flexibility, and working memory. Despite a growing need for evidence-based assessments of EF for autism populations, statistical models of many commonly used measures of EF, including the Delis-Kaplan Executive Function System (D-KEFS), have not been investigated for a sample of autistic participants. The purpose of this study was to address a gap in the literature regarding the latent structure of the D-KEFS in a sample of autistic individuals. The D-KEFS is one of the most widely used clinical assessments of executive function, but its factor structure has not been examined in a sample of autistic participants. Reliability analyses were performed for sample subgroups based on participants' clinical and demographic characteristics, including IQ, autism severity, age, and race/ethnicity. Verbal Fluency (VF) was found to consistently decrease or not affect the overall reliability score. Additionally, one- and two-factor structure models were tested for the D-KEFS with a sample of autistic participants. The one-factor model was not found to be a good fit for the data. However, the two-factor model, with Cognitive Flexibility and Abstraction latent factors, was found to fit the data relatively well. This two-factor model was reexamined excluding the VF observed variable, resulting in a better overall model fit. Communication deficits are a common feature of autism, which explains why the VF task, that requires participants to produce novel words, may not be an adequate measure of executive function for autism populations.