Comment on “On the Power of the F-test for Hypotheses in a Linear Model” by Griffiths and Hill (2022)

Abstract

Griffiths and Hill (2022) showed that when testing an hypothesis in linear models, it can sometimes be advantageous to incorporate components into the hypothesis that are true. They assume a true linear model and that a constraint R 2 β = d 2 is true. They are interested in comparing the powers of three F tests for H 0 : R 1 β = d 1 :