Saturday, November 10, 2012

You really should be reading Andrew Gelman's blog

I considered this point (made by Andrew Gelman) to be really interesting:
All statisticians use prior information in their statistical analysis. Non-Bayesians express their prior information not through a probability distribution on parameters but rather through their choice of methods. I think this non-Bayesian attitude is too restrictive, but in this case a small amount of reflection would reveal the inappropriateness of this procedure for this example.
I had never phrased things like this but it does seem to be a sensible description of what I (essentially a Frequentist) actually do in real life.  This may be a very good teaching point and it nicely illustrates that all analysis involves priors -- just that some are more explicit than others.

No comments:

Post a Comment