We then identified the shape of externalizing trajectories within each of the three sub-samples (for further details, see Hussong, Wirth et al., 2007). First, we plotted the means of Item Response Theory scores across age and examined iterative analyses to examine the optimal functional form of the resulting trajectories (i.e., linear, quadratic and piece-wise liner) following the strategies described by Bollen and Curran (2006). Intercepts in all models were centered at age 13. The nesting of repeated observations in multiple siblings within family resulted in 3-level models in the MLS. In the AFDP, we used a 2-level model due to the more simple nesting of repeated observations in target participants (who were independent of family). All trajectory models were estimated separately for each reporter using restricted maximum likelihood as implemented in SAS’s MIXED procedure (Littell, Milliken, Stroup, & Wolfinger, 1996) following strategies described in Singer and Willett (2003).
