Mr. Kim et al. Reply

Abstract

We write to respond to an erudite letter to the editor by Zachary J. Williams.¹ that extended analyses presented in our recent publication² investigating the latent structure of autism phenotypic symptoms. The letter by Williams reported results of a bifactor analysis using the standardized loadings and interfactor correlations derived from a Schmid-Leiman orthogonalization of the optimal correlated factor model that we identified. Williams examined the relative contributions of the general and three specific autism symptom factors (ie, social interaction [SI], communication [COM], and restrictive and repetitive behavior [RRB]) to the autism symptom indicators. Those analyses indicated that the majority of autism spectrum disorder (ASD) symptom covariation could be explained by a single general autism factor, whereas the three specific autism symptom factors accounted for minimal covariation among indicators. To pursue the intriguing points raised in the letter, we performed additional analyses using the original raw data.