Interesting article: The abuse of power: the pervasive fallacy of power calculations for data analysis. Hoenig JM and Heisey DM. The American Statistician 2001; 55(1): 19-24. Description: A technical reference on why post hoc power is always bad, even when you use a clinically relevant estimate for effect size. The article coins the acronym PAP (power approach paradox) to describe the situation that "higher observed power does not imply stronger evidence for a null hypothesis that is not rejected." The authors also point out that once a confidence interval is calculated, the post hoc power calculation adds nothing further to the interpretation. The description of this article was written on 2007-11-27, 2008-01-12. Category: Interesting articles, Category: Post hoc power