http://www.gemini-grp.com/Bayes/BerryNatureArticle.pdf

1. Berry claims that Bayesian clinical research allows for adaptive research designs, including the possibility of stopping clinical studies early, without sacrificing the analyst's ability to draw valid and useful statistical conclusions from them. Why is this possible?
2. Discuss at least one other important way in which Bayesian statistics out-performs frequentist statistics in clinical contexts, according to Berry
3. Berry also notes a handful of important limitations that the Bayesian approach faces, in clinical applications and more generally. Discuss at least one of these