Abstract
Trajectory modeling has recently come into favor in criminological research. Advances in methodology and statistics have provided the opportunity for criminologists to study, document, and understand developmental trajectories of criminal activity and behavior. Through the use of longitudinal data, criminologists can study behavioral change and the change in patterns of criminal activity over the life course, with such inquiries as: Are there certain trajectories of offending that begin early, continue through adolescence, and persist into adulthood? Are there certain offending trajectories that can be considered late onset? Are there certain risk and protective factors that differentially predict which trajectory an individual’s offending pattern may follow? And, do these offending trajectories generalize across biological sex, race, ethnicity, social class, time, and geographic location? Numerous longitudinal studies have been conducted throughout the world, examining these various research questions. The use of longitudinal data and trajectory models has provided criminologists with rich data to analyze research questions that would not have been possible to address through cross-sectional research. Historically, several findings exist that are common to longitudinal studies: misbehavior starts early in life and is often identifiable early, but not always. The majority of longitudinal studies have charted persistence of criminal behavior effectively, but they have not been as successful in charting desistance. Longitudinal studies have, for the most part, found the correlates of persistence and desistance of delinquent behavior to be dissimilar and not necessarily indicative of factors that contribute to early onset. And advances in statistical tools have allowed researchers to examine behavioral patterns of criminal activity more directly. A comprehensive examination of the various trajectory studies that have been conducted recently by sample type and time period as well as a summary of the key findings from these various trajectory studies will give further insight into what the field of criminology has learned from this trajectory methodology. Some of the key information discovered from applications of the trajectory methodology in criminology includes determining that adolescent-limited and life-course-persistent trajectory patterns can be identified. On average, three to five trajectory groups are identifiable using the trajectory methodology, and the relative number of trajectory groups often varies as a function of sample size. There are a variety of directions that can be taken in regard to longitudinal research and the trajectory methodology to improve the extant knowledge base in the field of criminology. For example, there is a need to further disaggregate the general offending trajectories across sex, race, ethnicity, and cultural contexts. Also, more literature is needed on the identification of multiple risk and protective factor domains such as family, peers, and community context that may serve as important mechanisms for differentiating offender trajectories. Finally, it is important for future criminological research using the trajectory methodology to extend the observation period across multiple developmental periods when data permit, in order to more fully capture the development of offending over the life course.