Latent growth curve modeling as an integrative approach to the analysis of change

Latent Growth Curve Models (LGCM) are discussed as a general data-analytic approach to the analysis of change. Conventional, but popular, methods of analyzing change over time, such as the paired t-test, repeated measures ANOVA, or MANOVA, have a tradition, which is quite different from the more recently developed latent growth curve models. While the former originated from the idea of variance decomposition, the latter have a factor analytic background. Accordingly, “traditional methods”, which focus on mean changes, and “new methods”, with their emphasis on individual trajectories, are often treated as two entirely different ways of analyzing change. In this article, an integrative perspective is presented by demonstrating that the two approaches are essentially identical. More precisely, it will be shown that the paired t-test, repeated measures ANOVA, and MANOVA are all special cases of the more general latent growth curve approach. Model differences reflect the underlying assumptions, and differences in results are a function of the degree to which the assumptions are appropriate for a given set of data. Theoretical and practical implications are set forth, and advantages of recognizing latent growth curve models as a general data-analytic system for repeated measures designs are discussed.