This month our very own Chris Weaver will present on “Use of Bayesian statistics in Science: Advantages and Pitfalls” (see below for more detail), and then we’ll have a discussion.
Here’s a little bit more on what he’ll be covering:
Bayesian techniques are popular in the rationality community, and also in science. The goal of this talk is to explore somewhat the ways in which they are and aren’t useful for determining and communicating scientific results. The subject is necessarily mathematical, but to the extent possible will be covered without making the audience know or delve deeply into a lot of math.
Much of the inspiration for this topic is drawn from the classic paper by Robert Cousins, “Why isn’t every physicist a Bayesian?”, a preprint version of which can be found at https://cds.cern.ch/record/274483/files/SCAN-9501056.pdf. This is not required reading, just background material for anyone who wants to dig in further!
Another interesting resource, for those who haven’t already seen it, is https://plato.stanford.edu/entries/probability-interpret, which may also be something to look at in this context, but is again by no means required.