Recent evidence has suggested that different ODD behavioral dimensions may have different correlates and outcomes (Burke, in press; Stringaris & Goodman, 2009a, 2009b). We examined whether there was empirical support for sub-factoring ODD symptoms that may in turn have implications for understanding links with CD. Our findings supported a two factor model with close similarities to the dimensions proposed by Stringaris and Goodman (2009b) on a priori theoretical grounds. Our irritability factor contained the same items as their irritability dimension, and our headstrong factor also contained all the items they proposed as indexing ‘headstrongness’. In addition, however, our headstrong factor also included the spiteful/vindictive symptom that Stringaris and Goodman treat as a separate dimension. In the GSMS spiteful/vindictiveness is indexed by a single symptom (following the DSM-IV specification), whereas in Stringaris and Goodman’s work separate items index spitefulness and vindictiveness. It is possible that a separate third factor would have been identified here if that approach had been taken. Future studies would benefit from including a greater range of items designed to measure each hypothesized dimension. Our factor analytic results
