To examine whether a bifactor model with a general ADHD factor and domain specific factors of inattention, hyperactivity and impulsivity was supported in a large general population sample of children. We also explored the utility of forming subscales based on the domain-specific factors.
Child mental health questionnaires were completed by both teachers and parents of all children in grades 2-4 in Bergen, Norway. Confirmatory factor analysis was performed on the ADHD items of a modified version of the Swanson, Nolan and Pelham Questionnaire-IV (SNAP-IV) for 6,237 children.
The bifactor model showed very good model fit with a strong general ADHD factor and specific factors for impulsivity and inattention. The subfactors, especially hyperactivity, generated from the SNAP-IV ADHD items conveyed little unique variance in the model.
The findings in this general population sample with a strong general ADHD factor in the bifactor model supports the view on ADHD as a unitary concept with specific domain factors for inattention and impulsivity, but not for hyperactivity. The bifactor model questions the utility of constructing ADHD subscales by the use of a simple sum score when using the SNAP-IV.