ABSTRACT
This chapter introduces the concepts of Bayesian inference and advances the argument concerning the alignment of Bayesian approaches to statistical inference. The authors develop ideas and practices associated with exchangeability one of the foundational elements in Bayesian model conceptualization and construction. The chapter provides a high-level contrast between the Bayesian and frequentist approaches to inference. It also serves to further characterize features of Bayesian approaches that are distinct from frequentist approaches, and may assuage concerns of those new to Bayesian inference. The distinctive feature of Bayesian approaches is the treatment of parameters, indeed, all entities, as random, with accompanying distributional specifications. The role of the prior distribution is often the focus of criticism from those skeptical of Bayesian inference, but the aforementioned issues of justification and examination apply to other features of the model, including the specification of the likelihood.
