Sir.—The pervasive confusion among medical professionals regarding statistical concepts and their proper use has been noted in several recent reports.1-3 Expository articles written to rectify this problem are thus important additions to the medical literature. There is an inherent danger, however, that such an effort can create additional misconceptions.
A case in point is the generally well-written article "Standard Deviation, Standard Error: Which 'Standard' Should We Use?" by Dr Brown (Journal 1982;136:937-941). While Dr Brown's presentation is basically correct, several nagging technical errors in the article could potentially lead to further confusion. As Dr Brown states on page 941 of his article, "When one sees medical publications with inappropriate, confusing, or wrong statistical presentation, one should write to the editors."
First, the median is that number compared to which half the data values are larger and half are smaller. The definition given in Dr Brown's article (p 937)