ABSTRACT
How does the design of a sample survey, an experiment, or an observational study affect the models that we construct for a Bayesian analysis? How does one analyze data from a survey that is not a simple random sample? How does one analyze data from hierarchical experimental designs such as randomized blocks and Latin squares, or nonrandomly generated data in an observational study? If we know that a design is randomized, how does that affect our Bayesian inference? In this chapter, we address these questions by showing how relevant features of data collection are incorporated in the process of full probability modeling.
