Latent Trajectory/Growth Curve Analysis: A Structural Equation Modeling Approach

Instructor(s): Ken A. Bollen, Department of Sociology, University of North Carolina at Chapel Hill

A powerful method for analyzing longitudinal data is Latent Trajectory Analysis (LTA). LTA allows each case in a sample to have individual trajectories ("latent curves" or "growth curves") representing change over time. In addition to mapping these trajectories, LTA allows researchers to examine the determinants of these trajectories or to relate the trajectories of one variable with those of another. The approach to LTA in this course draws on the strengths of structural equation modeling (SEM), and the primary goal is to introduce participants to the theory and application of LTA. The course begins with a conceptual introduction to LTA, a description of research questions that are well-suited for the technique, and a review of SEMs. The remainder of the course will cover the following topics: LTA models for a single variable with and without predictors of differences in trajectories; modeling nonlinear trajectories; the LTA model for multiple variables; the relation between the parameters governing the trajectories in two or more variables; incorporating predictors of multiple trajectories; and extensions to the LTA model. Participants should have prior training and experience with structural equation modeling and related software.