What does it mean to practice Bayesian statistics?
Andrew just posted Philosophy and the practice of Bayesian statistics, which links to an article he and Cosma Rohilla Shalizi wrote by the same name. He writes that it "... is flat-out entertaining to read. It's about philosophy, so it's supposed to be entertaining, in any case." While it's definitely not an elevator speech, I did find it entertaining to read, and I found that I agreed with most of what I read (or at least most of what I understood).
That may get the statisticians among us to consider reading the article, but I'd also encourage the system dynamicists among us to do the same. Here are a few tidbits to get you thinking and to encourage you to read it now:
"Social-scienti c data analysis is especially salient for our purposes because there is general agreement that, in this domain, all models in use are wrong -- not merely falsi able, but actually false. With enough data -- and often only a fairly moderate amount -- any analyst could reject any model now in use to any desired level of con dence. Model fi tting is nonetheless a valuable activity, and indeed the crux of data analysis."
"The data-analysis process -- Bayesian or otherwise -- does not end with calculating parameter estimates or posterior distribution. Rather, the model can then be checked, by comparing the implications of the fi tted model to the empirical evidence."
"In our view, a key part of Bayesian data analysis is model checking, ...."
"We are not interested in falsifying our model for its own sake -- among other things, having built it ourselves, we know all the shortcuts taken in doing so, and can already be morally certain it is false."
I could quote more, but I'll leave you with the link to the article itself in the hopes you, too, will read it and see why they do seek to falsify their models. Perhaps that will help us all benefit from an increased understanding of the role of models in the work we do.