A call for greater power in an era of publishing negative results

Anna L Oberhofer, Robert P Lennon

Abstract


Historically, medical journals published only positive results from experimentation.
With the growth of electronic media, many journals are now able to publish negative results as well. Because most medical experiments are designed to have different thresholds of significance for positive and negative results, this can lead readers to misunderstand the level of significance of a published negative result. We propose a technical shift – setting equal to β – to avoid this potential for misunderstanding.

Keywords


Statistics; Beta statistic; Power negative results

Full Text:

PDF


DOI: http://dx.doi.org/10.5644/ama2006-124.118

Refbacks

  • There are currently no refbacks.


2016 Department of Medical Sciences, Academy of Sciences and Arts of Bosnia and Herzegovina, Sarajevo, Bosnia and Herzegovina. All rights reserved.
 
The full text of articles published in this journal can be used free of charge for personal and educational purposes while respecting authors and publishers’ copyrights. For commercial purposes no part of this journal may be reproduced without the written permission of the publisher.