A Note on Misconceptions Concerning Prospective and Retrospective Power

• Published in: Journal of the Royal Statistical Society. Series D (The Statistician) Vol. 47; no. 2; pp. 385 - 388
• Main Authors: Zumbo, Bruno D., Hubley, Anita M.
• Format: Journal Article
• Language: English
• Published: Oxford, UK and Boston, USA Blackwell Publishers Ltd 01-07-1998; Blackwell Publishers
• Subjects: Conditional probabilities; Data analysis; Effect size; Misconception; Null hypothesis; Observational research; Population mean; Probabilities; Research design; Sample size; Statistical power; Statistical tests; Statistics
• ISSN: 0039-0526; 1467-9884
• DOI: 10.1111/1467-9884.00139

This paper addresses some misconceptions concerning prospective and retrospective power when journal editors, reviewers and readers want to know the observed power of statistical tests within a completed study. In addition, a practical solution to both the problem of computing the power after a study has been conducted and the critical evaluation of null findings is suggested.