ABSTRACT
This chapter begins with psychometric modeling paradigms with classical test theory (CTT). The CTT model has been conceived of in several distinct ways, differing in terms of whether an author develops the model from single or multiple observables, and from one of several different distributional notions of error. In CTT, the inferential target is the examinee's true score, so it is worthwhile to consider the various options in estimating an examinee's true score. The chapter discusses the development of CTT by considering a single observable assuming that the measurement model parameters and the hyperparameters are known. It provides the case where there are multiple observables, still assuming that the measurement model parameters and hyperparameters are known. The chapter presents the case where the measurement model parameters and hyperparameters are unknown. It also details a Bayesian approach to analysis in CTT, the first of several psychometric modeling paradigms to be considered.
