Comparing the external validity of the DSM-IV and Two-Factor models of childhood behavior problems

This study of 132 clinic-referred 6- to 12-year old boys with behavior problems examined the concurrent external validity of dimensions of the DSM-IV and Two-Factor diagnostic models. Both models used parent responses to the same DSM structured interview, focusing the study on conceptual differences between the models. Age at evaluation accounted for major variation and was held constant in all analyses. Both models, in both two- and three-dimension versions, had considerable external validity. The Two-Factor inattention-overactivity (IO) dimension predicted more external variables than the DSM-IV inattention (IN) dimension. DSM oppositional/conduct (OD/CD) accounted for some variation that is usually misattributed to DSM-IV hyperactivity-impulsivity (HI). In hierarchical analyses, the IO and AG dimensions of the Two-Factor model predicted more variation in more dependent variables than did the IN and HI dimensions of the DSM model of ADHD. However, when the often-ignored third dimension (OD/CD) was included, the complete three-dimension DSM-IV model accounted for more variation in more dependent variables than did the Two-Factor model. When the excluded third dimension (HI) was added to the Two-Factor model, the two three-dimension models were essentially equal in external validity.