Issues in reporting epidemiological studies: Can't we all agree on confidence intervals?BMJ 2005; 330 doi: https://doi.org/10.1136/bmj.330.7483.146-a (Published 13 January 2005) Cite this as: BMJ 2005;330:146
- Matthew W Gillman, associate professor (Matthew_gillman@hms.harvard.edu)
- Department of Ambulatory Care and Prevention, Harvard Medical School, Boston, MA 02215, USA
EDITOR—Pocock et al discuss many valuable caveats regarding published epidemiological analyses.1 The authors, however, perpetuate two misconceptions.
The first is that post hoc power calculations help the reader interpret already collected data. Power …