Taught by Gregory Hancock, Ph.D. 

Longitudinal data are ubiquitous throughout the social and behavioral sciences and beyond, where researchers have questions about the nature of change over time as well as its determinants. This two-day seminar provides a thorough introduction to latent growth curve models, which facilitate an assessment of longitudinal change from within the structural equation modeling (SEM) framework. 

The seminar will start with a quick review of SEM with measured and latent variables, illustrating the use of Mplus for such models. Next, latent means models, which add a mean structure to typical covariance-based structural models, will be introduced and illustrated with Mplus. The seminar will then review more traditional longitudinal models within an SEM framework (repeated measure models, panel models, etc.) to finish laying the necessary foundations.

The seminar will then move into a thorough coverage of traditional linear latent growth models, including but not limited to different time centering, uneven and varied time points, and time-independent covariates. Then topics will transition into more complex modeling variations, drawing from the following areas as time allows:

• nonlinear models
• spline models
• time-dependent covariates
• growth models for treatments and interventions
• multidomain models
• cohort-sequential models for planned missing data
• second-order growth models
• latent-difference score models
• growth models with categorical data
• growth mixture models
• power analysis in latent growth models