Bayesian model choice for the Poisson model « Robin Ryder's blog. Following Arthur Charpentier‘s example, I am going to try to post occasionally on material covered during my courses, in the hope that it might be useful to my students, but also to others.
In the second practical of the Bayesian Case Studies course, we looked at Bayesian model choice and basic Monte Carlo methods, looking at data about the number of oral questions asked by French deputies (Members of Parliament). We wanted to know whether male and female deputies behave differently in this matter. If we model the number of questions asked by a deputy by a Poisson random variable and denote their gender (1 for men, 2 for women), we thus need to choose between the following two models. History and Rationality Lecture Series - Daniel Kahneman.
Thinking, Fast and Slow: Daniel Kahneman: 9780374275631: Amazon.com.