What is the power of the study to refute the null hypothesis (zero anatomical difference between the case and control populations) when it is not true? Unfortunately, few published imaging studies consider this question explicitly. In general, the power of a study is proportional to the sample size (the number of subjects scanned), the effect size (the anatomical difference between populations), and the probability threshold or p value adopted for hypothesis testing. The p value will often be decided in relation to the number of tests conducted—the greater the number of tests, the smaller is the appropriate p value. Therefore the risk of low power and associated type 2 error is likely to be greatest when differences between two small groups have een multiply tested on the basis of many anatomical variables. In general, it is reasonable to be sceptical about negative findings reported in samples comprising less than 30 or 40 individuals per group.
Was this article helpful?
Put an end to the disappointment, frustration and embarrassment of... Erectile Dysfunction. Have Sex Like You are 18 Years Old Again. Have a natural powerfully stiff penis. Eject volumes of semen. Do it again and again night after night. Never make another excuse for not being able to get it up.