This Article  • References  • Full text PDF  • Send to a friend  • Save in My Folder  • Save to citation manager  • Permissions  Citing Articles  • Contact me when this article is cited  Related Content  • Similar articles in this journal  Social Bookmarking   What's this?

THOMAS B. NEWMAN, MD, MPH
Division of General Pediatrics Department of Pediatrics School of Medicine University of California 74 New Montgomery St, Suite 600 San Francisco, CA 94105

WARREN S. BROWNER, MD, MPH
Division of General Internal Medicine Department of Medicine School of Medicine University of California 74 New Montgomery St, Suite 600 San Francisco, CA 94105

Am J Dis Child. 1991;145(3):250.

	Since this article does not have an abstract, we have provided the first 150 words of the full text PDF and any section headings.

Sir.—We would like to comment on the multiple-comparisons problem posed by Brown¹ in his recent contribution about P values. The problem is: why should an investigator who wishes to compare treatments A and B be required to demonstrate a greater level of statistical significance if he or she studies treatment C at the same time? On the one hand, there is an apparent need to protect against the increased risk of a type I error associated with making multiple comparisons. On the other hand, it is counterintuitive that the conclusion for the comparison between A and B differs depending on whether the investigator had the resources to study treatment C at the same time.

Brown suggested leaving this enigma to experts in probability. While we do not claim such expertise, we have offered an explanation of this apparent contradiction that some of our colleagues have found helpful.² . . . [Full Text PDF of this Article]

CiteULike    Connotea    Del.icio.us    Digg    Reddit    Technorati    Twitter     What's this?