Confirmatory Factor Analysis of Oppositional Defiant Disorder within Clinically Referred Youth
Abstract
Each new edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5; APA, 2013) has been met with substantial criticism. Particularly, in DSM-5, two disorders were defined by very similar criteria. Oppositional defiant disorder (ODD) was defined as consisting of three dimensions - irritability, noncompliance, and spiteful/vindictive. Additionally, ODD has duration criteria that indicate its symptoms must be present for at least 6 months suggesting the presence of chronic irritability. DSM-5 also included disruptive mood dysregulation disorder (DMDD) as a disorder marked by the presence of chronic irritability in childhood and adolescence. The question of whether chronic irritability (i.e., DMDD) can be separated from ODD in clinical settings is a substantial question. Most studies indicate that DMDD and ODD have significant overlap (Freeman et al., 2016; Mayes et al., 2016). An alternate method is to examine whether ODD consists of independent or correlated dimensions. The factor structure of ODD can inform questions regarding whether irritability is a distinct dimension within ODD. Therefore, examining competing models of the factor structure of ODD in a clinical sample and externally validating the resulting dimensions should inform whether irritability should be treated as a unique, separate dimension of psychopathology or whether it is subsumed within a broader disruptive behavior dimension. The current study hypothesized that across parent and clinician ratings, ODD would have a multidimensional factor structure consisting of at least irritability and noncompliance factors. ODD’s factor structure would be best explained via a general ODD factor and two specific factors representing irritability and noncompliance. Additionally, the current study hypothesized that irritability and noncompliance will be differentially associated with internalizing symptoms and psychopathology and externalizing symptoms and psychopathology respectively. The hypotheses were partially supported. Implications for clinical decision making are discussed.