Network Psychometrics

Abstract

This chapter demonstrates how the Ising model can be estimated. It shows that the Ising model is equivalent to, or closely related to, prominent modeling techniques in psychometrics. The chapter introduces the general class of models used in network analysis called Markov random fields. It also shows that the Ising model is closely related to the modeling framework of item response theory (IRT), which is of central importance to psychometrics. The chapter illustrates how network models could be useful in the conceptualization of psychometric data. Latent variables have played a central role in psychometric models. As psychometrics starts to deal with network models, the Ising model offers a canonical form for network psychometrics, because it deals with binary data and is equivalent to well‐known models from IRT. The chapter focuses on the equivalence between the Ising model and multidimensional item response theory.