Applied Statistical Power Analysis and the Interpretation of Nonsignificant Results by Research Consumers

The publication of research findings that are not statistically significant presents a novel problem in interpretation of research results. The contribution of nonsignificant results depends in part on whether the statistical test was powerful enough to detect an effect of "meaningful" size. The primary responsibility rests with the authors of articles reporting nonsignificant results to demonstrate the worth of the results by discussing the power of the tests. If they do not assume this responsibility, then consumers of research should be prepared to conduct their own power analyses to aid interpretation of the research results. This article demonstrates the use of power analysis for the interpretation of nonsignificant findings. The power of many common statistical tests can be determined without difficult computation using Cohen's (1977) or Stevens's (1980) tables.