Bayesian networks (BNs) model multivariate distributions of discrete variables. When brought to bear in psychometric contexts they are typically constructed as modeling discrete observables as dependent on discrete latent variables. As such they are similar to the LCA models discussed in Chapter 13. There, we focused on models with a single latent variable. In this chapter, we focus mainly on models with multiple latent variables. In this light, BNs may be seen as generalizations of LCA models as developed in Chapter 13. In another light, BNs may be seen as instances of LCA models more broadly conceived, as models with multiple discrete latent variables may be recast as having a single discrete latent variable (see Section ...
Get Bayesian Psychometric Modeling now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.
