Investigating Stability and Change in Substance Use and Criminal Activity Using a Synthesized Longitudinal Modeling Approach 1

Abstract

The application of career perspectives in substance use and crime research has emphasized both stability and change. Curran and Bollen (2001) recently developed an analytic approach, the autoregressive latent trajectory model (ALT), which permits the analysis of both persistent individual differences in behavioral trends and stage-specific direct relationships. The present study utilizes this modeling approach with a sample of 524 offenders released from the California Youth Authority followed over seven years in early adulthood. The results of the analytic process suggest that the ALT model fit the data well. Comparatively, that model fit better than the autoregressive or LGC specifications and provides more information. The ALT model provides estimates that are relevant in understanding key parameters of the substance use and criminal career frameworks—both singly and jointly. The results of this investigation are then considered from the standpoint of future methodological development and substantive theoretical considerations of longitudinal patterns of substance use and crime.