A Bayesian wants everybody else to be non-Bayesian

Bayesian inference proceeds by taking the likelihoods from different data sources and then combining them with a prior distribution (or, more generally, a hierarchical model). The likelihood is key. For example, in a meta-analysis (such as the three examples in Chapter 5 of our book), you need the likelihood for each separate experiment. No funny stuff, no posterior distributions, just the likelihood. In a linear model setting, it’s convenient to have unbiased estimates. I don’t want everybody coming to me with their posterior distribution–i’d just have to divide away their prior distributions before getting to my own analysis.

Sort of like a trial, where the judge wants to hear what everybody saw–not their individual inferences, but their raw data. Anyway, it’s kind of funny since we’re always saying how Bayesian inference is the best, but really we don’t want other people preprocessing their data in this way. When combining subjective estimates, the challenge is that there are no pure, unbiased data points.

See part 1 of this talk for more details.

2 thoughts on “A Bayesian wants everybody else to be non-Bayesian

  1. Would a meta-analysis based purely on Bayesian posteriors give you a rational summary of other peoples' beliefs: a consensus of experts, if you will. Wouldn't this then mean that such a meta-analysis was answering a slightly different question?

    If so, I'll leave it to others to decide if the question it's answering is useful, and then I'll try and make a summary of other peoples' opinions. :-)

    Bob

  2. Found this well aged posting as I have been thinking about evaluating, contrasting and combining priors from different sources for an upcoming SAMSI program on meta-analysis

    http://www.samsi.info/programs/2008meta-analysisp

    Andrew's comment "When combining subjective estimates, the challenge is that there are no pure, unbiased data points" has me worried that it may be a very difficult challenge or least not the sort of usual statistical challenge.

    Did not seem to start much discussion either(at least 2 years ago)

    Keith

Comments are closed.