CitationSavitz, David A. & Olshan, Andrew F. (1995). Multiple Comparisons and Related Issues in the Interpretation of Epidemiologic Data. American Journal of Epidemiology, 142(9), 904-8.
AbstractEpidemiologists have an abundance of valid reasons to be self-critical and cautious in the interpretation of their results, usually presented in the discussion section of manuscripts. However, epidemiologists themselves and, to a much greater extent, nonepidemiologists often question the validity of a particular finding because of its context rather than because of the methods that generated it. In this commentary, we argue that a concern with multiple comparisons is unwarranted. A closely related theme is the concern with the investigator's perspective—how and when the idea for collecting and analyzing data occurred, which we also argue is irrelevant to assessing the validity of the product. Focus on such issues may lead to unjustified dismissal of meaningful results or exaggerated confidence in weak results and distract readers from focusing on the real threats to validity.
Reference TypeJournal Article
Journal TitleAmerican Journal of Epidemiology
Author(s)Savitz, David A.
Olshan, Andrew F.