GEORGE A. DIAMOND, M.D.; JAMES S. FORRESTER, M.D.
To the editor: Ellison and Seney (1) (along with five others who corresponded with us directly) correctly detected an error in our article on clinical trials (2) in panel A of Figure 1. The posterior probability curves, although properly calculated by a computer, were improperly labeled by a mortal—a pernicious demonstration of Murphy's law. The labels should be identical to those in panel B. We apologize for the error. Our conclusion—that a statistically significant result may not support the truth of a tested hypothesis when sample size is small—is still supported, however, by reference to panel B. The statement by
DIAMOND GA, FORRESTER JS. Posterior Probability in Clinical Trials. Ann Intern Med. 1984;100:457. doi: 10.7326/0003-4819-100-3-457_2
Download citation file:
Published: Ann Intern Med. 1984;100(3):457.
Copyright © 2018 American College of Physicians. All Rights Reserved.
Print ISSN: 0003-4819 | Online ISSN: 1539-3704
Conditions of Use