gwern comments on Beyond Bayesians and Frequentists - Less Wrong

36 Post author: jsteinhardt 31 October 2012 07:03AM

You are viewing a comment permalink. View the original post to see all comments and the full post content.

Comments (51)

You are viewing a single comment's thread.

Comment author: gwern 25 December 2012 10:21:14PM 3 points [-]

As Andrew Critch [6] insightfully points out, the Bayesians vs. frequentists debate is really three debates at once, centering around one or more of the following arguments:...

I think there's a much more important and fundamental debate you're missing in your taxonomy, and one of the wellsprings of LW criticism: the sub-category of frequentist techniques called null hypothesis testing. There are legitimate & powerful frequentist criticisms of NHST, and these are accepted and echoed as major arguments by many who are otherwise on an opposite side for one of those other debates.

For my part, I'm sure that NHST is badly misleading and wrong, but I'm not so sure that I can tar all the other frequentist techniques with the same brush.