So, far we have learned how to use Bayesian methods to address questions commonly asked by researchers—differences among groups and prediction. In Chapter 8 , we showed how we can compare groups on non-normal outcomes (Poisson and binomial distributed outcomes). The Bayesian approach to these problems only requires changing the likelihood—otherwise, the MCMC methods for sampling from the posterior remain the same. This consistency makes jumping from one kind of problem—comparing two groups on a normally distributed outcome—to another—comparing two groups on a binomial outcome—fairly simple. In this chapter, we continue working with non-normal outcomes, extending regression to accommodate binary outcomes.