Bayesian analysis for sociologists - An introduction

This article provides an applied introduction to Bayesian statistics for so ciologists. Unlike frequentist statistics, which attaches repeated-sampling frequencies to test statistics, Bayesian statistics directly describes unc ertainty about unknown statistical parameters with a probability distributi on With this foundation, much of Bayesian statistics follows from basic rul es of probability theory Three areas of Bayesian statistics are especially relevant for sociologists. First, hierarchical regression models allow seve ral levels of uncertainty into an analysis. Second Bayes factors provide a useful approach to the problems of model selection, model averaging, and po sterior inference about model indexes. Third, recent breakthroughs in estim ation methods offer valuable new tools for analysis of Bayesian models that were previously intractable.