Commentary on application of the bifactor S-1 model to multisource ratings of ADHD/ODD symptoms

Abstract

My commentary is organized into four sections. First, I summarize the reasons for the original interest in fitting symmetric bifactor models to ADHD data. Second, I summarize the concerns that Burns and colleagues raised with respect to fitting symmetric bifactor models to ADHD item-level data and describe their recommended alternative approach. Third, I raise two concerns that I had with their manuscript. Fourth, I conclude with a caveat and a general question about the merits of the continued study of the factor structure of ADHD symptoms.