significance test n.
In inferential statistics, any test designed to determine whether an alternative hypothesis achieved the required level of statistical significance to justify being accepted in preference to the null hypothesis. In a classical null hypothesis significance test, the null hypothesis is provisionally assumed to be true, and a calculation is then made of the probability of obtaining a result as extreme as the one actually obtained. If the probability is sufficiently small (by convention this significance level is often set at p < .05), the investigator is justified in rejecting the null hypothesis and therefore also in accepting its logical negation, namely the alternative hypothesis. The term ‘tests of significance’ was introduced by the English statistician and geneticist Ronald Aylmer Fisher (1890–1962) in 1925, on page 43 of his book Statistical Methods for Research Workers. Also called a statistical test.