ABSTRACT
At long last we now get serious about analyzing data. Chapter 1 introduced typical results from Bayesian data analysis. Chapter 2 covered fundamental scientific and mathematical ideas behind the Bayesian approach. Chapter 3 showed how to avoid the nasty calculus associated with Bayesian computations. Chapter 4 discussed a wide array of statistical concepts useful in Bayesian statistics. In particular, Chapters 2 and 4 discussed one sample data for Bernoullis, binomials, and normals with known precision. Example 3.1.3 also discussed two independent binomial samples. But all of those discussions focused on illustrating concepts of Bayesian statistics. Nowwe focus on analyzing data from one or two populations. Methods are presented for commonly used parametric models including the binomial, negative binomial, normal with unknown precision, and Poisson. In particular, Section 1 discusses inferences for proportions, rates and effect measures under binomial, negative binomial and case-control sampling. Section 2 discusses one-and two-sample normal populations. Section 3 discusses one-and two-sample Poisson data. All sections contain details of how to specify prior distributions. Material on analyzing one and two samples for time to event data (survival analysis/reliability analysis) appears in Chapter 12.
