Introduction

Many longitudinal studies of drug use began in the 1970s with cohorts of adolescents (Kandel, 1980; White, 1996). Participants in these studies are now adults, allowing researchers to identify the impact of adolescent drug use on adult role functioning and health. Recent work in the area of adolescent substance use and its consequences has begun to focus on developmental processes and recognizes the need to examine long-term outcomes beyond adolescence. Since the emergence of a renewed interest in a lifespan developmental perspective in the 1960s, views of human development have evolved considerably. Earlier descriptions of change patterns as universal, invariant, irreversible, and unidirectional have lost ground to the recognition that there is wide variability in human development throughout the entire life course. This recognition has contributed to a growing interest in individual differences in the paths and trajectories of human development from childhood to adulthood (Campos, Hinden, & Gerhardt, 1995; Magnusson, 1996).

The need to be able to differentiate between the contemporary and cumulative impact of drug use on developmental outcomes has led to the implementation of longitudinal designs that cover extended time periods and incorporate more than two measurement occasions. How ever, the techniques for analyzing longitudinal data have lagged behind (see Chapter 2 in this volume), and one of the foremost issues confronting researchers today is how to best analyze change (e.g., Collins & Horn, 1991).