Abstract
Developmental, life course (DLC) criminology has become a prominent paradigm in the study of crime, and with its emergence has come advances in estimation methods intended to answer research on questions of interest in this area (Piquero et al., 2003). For instance, researchers may be interested in the onset of offending, its continuance, and its cessation. Each of these questions can be considered in the context of more general longitudinal “growth” in offending behavior. Multiple alternative approaches to studying longitudinal trajectories of criminal behavior have been utilized. This study investigates three alternatives in order to extend recent work on the relative merits of these models for estimating criminal career trajectories. In particular, we examine model performance in two important cohort studies as a conduit to considering model results in terms of theoretical implications, subsequent refinement, and future research.