“Pour la Science” almost goes Bayesian!
After the strange views held on Bayesian statistics by the popular science magazine La Recherche, it is more than comforting to see the other popular science magazine Pour la Science to publish a more balanced paper on the role of statistical evidence, both frequentist and Bayesian. And by Andrew Gelman! This paper is actually a translation into French of a paper of Andrew with David Wiekliem, published earlier in American Scientist. I can only make one complaint about a missing reference to Laplace (the true father of Bayesian statistics!) who did study the difference between male and female births in his Théorie Analytique des Probabilités.