The Inexact Use of Fisher's Exact Test in Six Major Medical Journals
- W. Paul McKinney, MD;
- Mark J. Young, MD;
- Arthur Hartz, MD, PhD;
- Martha Bi-Fong Lee, PhD
Abstract
We reviewed the use of Fisher's Exact Test in 71 articles published between 1983 and 1987 in six medical journals. Thirty-three of 56 selected articles did not specify use of a one- or two-tailed test, and 12 (36%) of these actually used the one-tailed test. Five (42%) of these 12 articles contained at least one table in which the standard significance level of P<.05 was no longer met when a two-tailed analysis was run instead. Eleven (65%) of 17 articles with a statistical consultant compared with 10 (28%) of 36 articles without a consultant specified use of a one- or two-sided test. The use of Fisher's Exact Test without specification as a one- or two-tailed version may misrepresent the statistical significance of data. A uniform specification statement should be required for all manuscripts to correct such potential errors in interpretation.
(JAMA. 1989;261:3430-3433)
Footnotes
-
Reprint requests to the Department of Medicine, Box 135, Milwaukee County Medical Complex, 8700 W Wisconsin Ave, Milwaukee, WI 53226 (Dr McKinney).
