The p-value is, possibly, the most used tool in applied statistics, although it has some important shortcomings. The q-value was introduced by Storey (2003) as a byproduct of the positive false discovery rate. In this article, it is shown that the q-value could be considered as a good alternative to the p-value, because the q-value exhibits better properties than the p-value and, moreover, the q-value is easier to interpret than the p-value.

Keywords: Coherence, hypothesis testing, positive false discovery rate, p-value, q-value.

AMS Subject classifications: 62F03, 62F15.