To the Editor.
—The recent article by Drs Browner and Newman1 about the civil rights of P values is a useful and imaginative description of an important thought process in medical scientific inference. It is to be hoped that the article will achieve a wide readership.One aspect of the present publication is puzzling, however, to a person with an interest in the historical development of ideas in epidemiology and biostatistics. The puzzling aspect is the apparent need for the publication. Indeed, the essence of the argument was already known and was taught to physicians more than a half century ago. To substantiate this claim, we quote almost a full paragraph from the 1930s book entitled An Introduction to Medical Statistics by Woods and Russell,2 which was intended as a textbook for physicians who took graduate courses at the London School of Hygiene as organized by Professor M.