Chapter 8: Using Latent Growth Modeling to Evaluate Longitudinal Change

Researchers are frequently interested in understanding how some aspect of an individual changes over time. The focus of their investigation might be on general outcomes such as behavior, performance, or values; or it could be on more specific aspects such as substance abuse, depression, communication skills, attitudes toward disabled veterans, or advancement of math aptitude. Regardless of an investigation’s focus, the real attraction in longitudinal studies is in understanding how change comes about, how much change occurs, how the change process might differ across individuals, and what the determinants of that change are.

Various methods can be used to analyze longitudinal data (see, e.g., Collins & Sayer, 2001; Gottman, 1995). Among the more traditional methods is analysis of variance (ANOVA), multivariate analysis of variance (MANOVA), analysis of covariance (ANCOVA), multivariate analysis of covariance (MANCOVA), and auto-regressive and cross-lagged multiple regression. No one method is necessarily superior, but each has strengths and shortcomings that researchers should be aware of in order to select the analytic method best suited for the particular research context.