[Skip to Navigation]
October 1940


Author Affiliations


From the Division of Biostatistics, Department of Preventive Medicine and Public Health, University of Minnesota.

Arch Intern Med (Chic). 1940;66(4):848-850. doi:10.1001/archinte.1940.00190160065006

A discussion of the range of normal blood pressures has been presented recently by Robinson and Brucer.1 The authors have wisely emphasized the need for more accurate standards in judging the normality of any blood pressure level shown on clinical examination. In the paper of reference they essayed to provide such standards, based on "an exhaustive statistical study of 7,478 men and 3,405 women selected at random." Disregarding the possible clinical soundness of their deductions, it cannot be agreed that their method of analysis of the data is statistically sound. The study appears to have been characterized both by biased selection of the basic data and by application of frequency curve probablity values to distributions which certainly do not follow the function used. It is the purpose of this note merely to point out a few of the statistical errors that affect the validity of the conclusions. Unfortunately mere

Add or change institution