Syndicate content

Add new comment

Submitted by Steve Ziliak on

Lovely day for a regression! Gosset, and Guinness, are good. But Fisher's daughter Joan Fisher Box does not get the history right. Gosset's economic approach to balanced designs of experiments was shown to be more precise and profitable for the great brewery. Gosset rejected Fisher's qualitative approach to p-values and unrealistic assumptions about uncorrelated material and randomization.
https://www.aeaweb.org/articles?id=10.1257/jep.22.4.199