Multiple Indicator–Multiple Indicator Cause, Mixture, and Multilevel Models

In this chapter we continue with our presentation and discussion of SEM model applications. Specically, we present an example where latent variables are predicted by observed variables (MIMIC model); an example where continuous and categorical variables are included in the model (mixture model); and nally an example where nested design data occur (multilevel model). All three of these SEM applications are unique and are not possible using traditional statistics (analysis of variance, etc.).