The general outcome of a multiple comparisons.

A total of m null hypotheses are tested. FP is the number of Type I errors or the number of false positives (rejected true hypotheses). Physical significance as indicated in the first column means the existence of a real effect, whereas statistical significance refers to the detection of such effect by means of measurements. FN is the number of Type II errors or the number of false negatives (false hypotheses not rejected). The number R of rejected hypotheses is an observable random variable, while FP, FN, TP and TN are unobservable random variables. The number of true null hypotheses is also unknown in practice. The empirical type I error rate is defined by FP/, while the empirical type II error rate is defined by FN/ and the estimated average power is TP/. See [12] and [13].