Assigning values to latent variables

In this chapter we discuss methods for assigning values to latent variables for individual clusters. The clusters could for instance be subjects in measurement and longitudinal modeling or schools in multilevel modeling. For continuous latent variables we will refer to this as latent scoring (factor scoring or random effects scoring) and for discrete latent variables as classification. Note that this terminology does not distinguish between different response types; factor scoring would for instance include scoring for IRT models with dichotomous responses.