One advantage of a longitudinal mixture approach has over traditional growth curve analysis is that it allows for relations between personality trajectories and their etiologically relevant time-varying covariates, or dynamic covariates, to be more precisely explored in an empirical framework (Sher et al., 2004). Dynamic covariates of a given construct may exhibit a range of developmental patterns (see Nuechterlein and Dawson, 1984; Sher et al., 2004). For example, changes in alcohol use may exhibit patterns of covariation with impulsivity designated as a course tracker. Under this pattern of covariation, individuals that decreased in impulsivity would also exhibit corresponding decreases in alcohol use; whereas, individuals that failed to make normative decreases/made shallower decreases in impulsivity would not display reductions in alcohol use. Conversely, failures to make normative decreases in impulsivity may be more accurately depicted as developmental lag markers of alcohol use. This pattern of covariation would be reflected by slowed or inhibited normative developmental changes in alcohol use by individuals that exhibited shallower or nonnormative decreases in impulsivity. For example, under this pattern in the current dataset that spans ages
