ABSTRACT
For a long time research on typologies of delinquents was either based on pre-conceived notions with little empirical support (e.g., Lombroso, 1912) or was derived from cross-sectional data only (Clinard & Quinney, 1973; Roebuck, 1966). Delinquent types based on cross-sectional data have many disadvantages in that they fail to capture changes in offending that take place with age, and that differentially affect some youth more than others (Gibbons, 1975). With the increased availability of multiwave longitudinal data over the past decade using self-reports and official records, it has proven more feasible to investigate the dynamic aspects of developmental typologies of delinquents. Buttressed by major improvements in statistical software to analyze such developmental trajectories, a new generation of findings has emerged showing that some categories of delinquents develop offending in different ways from other categories of offenders (Barker et al., 2007; Lacourse et al., 2002; Nagin, Farrington, & Moffitt, 1995; Nagin & Land, 1993, White, Bates, & Buyske, 2001).
