07-modelSelection-00-intro.Rmd

# Bayesian Model Choice

In Section \@ref(sec:Bayes-multiple-regression) of Chapter 6, we provided a Bayesian inference analysis for kid's cognitive scores using multiple linear regression. We found that several credible intervals of the coefficients contain zero, suggesting that we could potentially simplify the model. 
In this chapter, we will discuss model selection, model uncertainty, and model averaging. Bayesian model selection is to pick variables for multiple linear regression based on Bayesian information criterion, or BIC. Later, we will also discuss other model selection methods, such as using Bayes factors.