ABSTRACT
This chapter develops the core ideas from probability-based reasoning for inference with measurement models and interprets them from a sociocognitive perspective. The ideas are as follows: (a) a subjectivist-Bayesian stance on probability, and how it is particularly suited to model-based reasoning in general and educational measurement models in particular; (b) conditional probability and conditional independence, and their “as if” role in framing the measurement-model narrative; (c) updating belief through Bayes theorem; (d) assembly of measurement models from probability-model fragments that correspond to narrative-theme building blocks; and (e) the situated nature of meaning of elements and probabilities in applications.
