The standard error of the mean (SEM) is an uncertainty measure on which both p-values and confidence intervals are based. It is, however, often mistaken as a measure of variation. To estimate how often this confusion is manifested in publications Ko et al. reviewed all original articles published in four selected journals during 2011.
The overall frequency with which SEM was inappropriately reported wa 13.6%. The misuse rate was lowest (2.6%) in Acta Obstetrica & Gynecologica Scandinavica, and highest (22.7%) in the American Journal of Obstetrics and Gynecology. The citation rate was, unfortunately, higher in articles misusing SEM. The misuse rate also correlated with citation half-life.
The authors conclude that inappropriate reporting of SEM is common, and that valuable articles may be in danger of being wasted by misuse of statistics.
1. Ko WR, Hung WT, Chang HC, Lin LY. Inappropriate use of standard error of the mean when reporting variability of study samples: A critical evaluation of four selected journals of obstetrics and gynecology. Taiwan J Obstet Gynecol 2014;53:26-29.