ABSTRACT
This chapter discusses the treatment of Bayesian psychometric modelling. Bayesian approaches to psychometric modeling serve us in two ways: they enable us to tell better stories, and provide us better ways to tell them. It summarizes the thrust behind adopting a Bayesian approach, which cuts across the modeling families and choices. Bayesian perspectives have also allowed for the resolution of paradoxes and deeper understandings of commonalities that cut across latent variable modeling families. The chapter shows that Bayesian approaches allow one to better build and reason with statistical models. Models reflect our beliefs and purposes, which may partially conflict. This necessitates careful thought and underscores the larger point that modeling is often an exercise in making explicit what are the relevant features of the situation to the inference at hand.
