Setting priors

Aram

Hi everyone,

I have read that we could use the default priors contained in the BayesFactor package. It worries me that I have read that using Bayesian inference with default priors, is like eating cake without sugar and flour (can't remember the metaphor tbh). Somehow they say it's pointless to do Bayesian inference without adjusting priors.

Also about the interpretation, I read on Richard's blog that the magnitude of the BF depends on what you were expecting. How can I get from previous papers on my topic, the effect size of my factors? He gives an example saying:

Is 10 kilometers a long way? It is if you're walking, it isn't if you've just started a flight to Australia.

I can see how this 10 is different in this example, but how can I know, from previous literature without Bayesian inference, which factors have big effects?

EJ

There's an entire statistical field called "objective Bayes", and their goal is to pick priors that fulfill a number of general desiderata and are fit for use across a wide range of situations. In my experience, within the limits of reason(!), the results do not differ dramatically if you use a subjective prior. For a concrete example see the blog post series on BayesianSpectacles.org. Next week's blog post will discuss the strength of the evidence and how to interpret it.

Cheers,
E.J.