Advocacy and Objectivity in Science

J. Scott Armstrong, University of Pennsylvania

Article comments

Postprint version. Published in Management Science, Volume 25, Issue 5, May 1979, pages 423-428.
Publisher URL: http://mansci.pubs.informs.org/

The author asserts his right to include this material in ScholarlyCommons@Penn.

Abstract

Three strategies for scientific research in management are examined: advocacy, induction, and multiple hypotheses. Advocacy of a single dominant hypothesis is efficient, but biased. Induction is not biased, but it is inefficient. The multiple hypotheses strategy seems to be both efficient and unbiased. Despite its apparent lack of objectivity, most management scientists use advocacy. For example, 2/3 of the papers published in a sampling of issues of Management Science (1955-1976) used advocacy. A review of the published empirical evidence indicates that advocacy reduces tire objectivity of the scientists. No evidence was found to suggest that this lack of objectivity could be overcome by a "marketplace for ideas" (i.e., publication for peer review). It is recommended that tire method of multiple hypotheses be used.