McNemar test

n. In statistics, a form of chi-square test for dichotomous variables which tests whether the two possibilities are equally likely while making no assumptions about the population distribution. This test is often used when testing the effectiveness of a treatment in eliminating a disease.