1. Issues of Power: Clarification.
- Author
-
Brewer, James K.
- Subjects
CRITICISM ,BAYESIAN analysis ,STATISTICAL decision making ,PROBABILITY theory ,RHETORIC ,PSYCHOLOGY - Abstract
The article comment on the criticisms by researchers J. Cohen C. M. Dayton and Donald L. Meyer on a survey paper by the author on the power of statistical tests. The major point of all the criticisms was on the supposed "after-the-fact" or post hoc usage of power. Two other issues dealing with the conditional nature of power and the reporting schemes are also mentioned. The author finds it difficult to be upset with the major points proposed by Meyer outside of misinterpretations. The author has no objections to the viewpoint presented in the paper. To some people the classical approach makes equally good sense, but the author chooses not to be pigeonholed as a "classicist" or a "Bayesian." The author believes a reasonable researcher can and should entertain all philosophical viewpoints, and select that one or several with which he is most comfortable. Once the Bayesian viewpoint is adopted, then the power issue is trivial. However, the Bayesian viewpoint has not been adopted and utilized for published research very often in the behavioral sciences.
- Published
- 1974
- Full Text
- View/download PDF