PATRICK A. HESSEL, Ph.D.
To the editor: Dr. Bailar (1) makes an important distinction between a priori and post hoc tests of hypotheses, stating that probability statements are only valid when hypotheses are "fully developed before the data are examined in any way." Although this distinction is extremely important, there is another kind of test that is very commonly done but does not easily fit into one of these two categories.
In many instances during the analysis of a study, a significant or nonsignificant result appears, whether expected or unexpected, whether suggested by the data or not. During deliberations about the real significance of
HESSEL PA. Science, Statistics, and Deception. Ann Intern Med. 1986;105:974. doi: 10.7326/0003-4819-105-6-974_1
Download citation file:
Published: Ann Intern Med. 1986;105(6):974.
Copyright © 2018 American College of Physicians. All Rights Reserved.
Print ISSN: 0003-4819 | Online ISSN: 1539-3704
Conditions of Use