developmentally relevant measures (trajectories) of externalizing behavior. Moreover, the GMM presented here extends that which was previously reported by our group in relation to GABRA2 and parental monitoring (Dick et al., 2009). Findings from the initial study indicated that a two-class solution best described the developmental heterogeneity in externalizing behavior between the ages of 12 and 22. In that study, however, trajectories were defined only in terms of linear growth. Given accumulating evidence that differential rates of change in externalizing behavior are likely to be observed over such an extended period of time (Bongers et al., 2004; Nagin & Tremblay, 1999), we opted for a less restrictive, and potentially more realistic model in the present analyses, allowing for the possibility of non-linear change via the inclusion of a quadratic growth parameter.
