Statistician mocks frequentists for patching one flaw and ignoring the rest

RexDouglass · x · 2026-09-29

The author mocks frequentist statistics: all effort goes to patching the single vulnerability of running 2 regressions and reporting 1, while everything else goes unexamined. He cites asking an author to demonstrate building a synthesized prior from the latest SOTA — refused, with only a year-old guess on stale info offered instead.

Original post →

More from Research

Research channel →