Structural Equation Modeling


This workshop will help participants develop skills in defining, estimating and testing graphical and latent variable models, and structural equation models in particular. After a general overview of concepts, regression approaches with observed variables will be demonstrated (e.g. path analysis). This will be followed by an overview of latent variable modeling techniques (‘factor analysis’), including categorical latent variables (mixture models). The second part of the course will focus on SEM generally, including model construction, interpretation, and issues in assessing such models. Specific time may be spent on a popular class of SEM, latent growth curve models, and time permitting, an overview of extensions and related techniques will be provided.