Trajectory Analysis of the Transition to Junior High School

Summary There are important limitations to analysis of variance (ANOVA) as a melhod for longitudinal prevention research. We consider in particular the limitations of ANOVA for identifying different courses of adjustment during life transitions and the proportion of the population who manifest each course. To complement traditional ANOVA, we argue for the use of cluster analytic techniques to uncover differences in longitudinal trajectory. This method is illustrated with two-year longitudinal data on self-esteem during the transition from elementary school to junior high school. Four contrasting trajectories are identified: Consistently High (35%), Chronically Low (13%), Steeply Declining (21%), and Small Increase (31%). The identification of the Steeply Declining trajectory is considered especially important for prevention. The trajectory approach yields distinctive information for identifying and understanding at-risk groups, for informing the design of prevention and policy options, and for increasing the specificity of evaluation efforts. Though not without limitations, the results are considered promising enough to warrant use of this approach for studying a wide array of outcomes and life transitions.