Quantitative biases analyses using Bayesian methods

This document was developped as part of a 5-day course on quantitative biases analyses developed with my colleague Ian Dohoo, from University of Prince Edward Island. In this course we proposed ways to adjust, during the statistical analyses phase, for the biases generated by the study design. In my part of the course, I reviewed the Bayesian methods that can be used to adjust for the different biases (selection and misclassification biases, or for an unmeasured confounder) directly in a regression model. It covers a wee bit of theory (though most of the theory was covered in the course), but this is mainly the R and R2OpenBUGS codes for conducting such analyses. There is a series of exercises (named Exercise 1 - Intro Bayesian to Exercise 6 - Multiple biases) that can be done; they can be downloaded from the course repository on my Github account. Special thanks to Henrik Stryhn, also from University of Prince Edward Island, who reviewed a previous version of this document.