LEONARD E. BRAITMAN, PH.D.
In this issue, Diamond and Forrester (1) use posterior probabilities to improve on traditional methods of eliminating random variation (chance) as a likely explanation of reported clinical trial results. Their bayesian analysis can be viewed as part of a general four-step approach to the evaluation of statistically significant (positive) findings of a randomized controlled trial:
1) elimination of chance as a likely explanation of the results,
2) examination of clinical significance of results using an appropriate measure of the strength of the statistically significant association,
3) identification of experimental biases in the trial (2, 3) and analysis of the possible
BRAITMAN LE. Statistical, Clinical, and Experimental Evidence in Randomized Controlled Trials. Ann Intern Med. 1983;98:407–408. doi: 10.7326/0003-4819-98-3-407
Download citation file:
Published: Ann Intern Med. 1983;98(3):407-408.
Copyright © 2018 American College of Physicians. All Rights Reserved.
Print ISSN: 0003-4819 | Online ISSN: 1539-3704
Conditions of Use