ABSTRACT
This chapter highlights the use of the kinds of models in situations where individuals have been randomly assigned to one or more treatment groups in a randomized experiment. It deals with structural equation modeling (SEM) analyses focused on isolating aspects of experimental manipulation that alter the dynamic processes, both within and between variables. The chapter presents contemporary statistical models useful for testing dynamic hypotheses using longitudinal experimental data. The dynamic SEM approach offers a practical approximation of dynamic interpretations for traditional experimental data with a standard collection of repeated measures. The chapter utilizes models with latent variables to represent group and individual changes in a bivariate dynamic system across different variables at different occasions of measurement. Most of the models are based on fitting observed raw-score longitudinal growth data to a theoretical model using maximum likelihood estimation.
