Statistical Power

Abstract The term “statistical power” is used in null hypothesis significance testing (NHST) to refer to the probability of rejecting the null hypothesis (thus providing support to the theory being tested) when the null hypothesis is false (i.e., the theory is true). Adequate power is having a large enough probability of rejecting the null hypothesis when the theory is true and of clinical and/or practical significance. Consideration of power in designing NHST studies prevents the waste of time, resources, and sometimes reputations, that accompany failed studies, those that neither support nor clearly refute the theory being tested, leaving the state of knowledge where it was before the study was done.