two-tailed test

n. In inferential statistics, a two-tailed (2-tailed) test is one that tests the null hypothesis that an observed result is less extreme in difference from the mean than would occur by chance. So a result either greater or lesser than the mean will be determined to be due to chance if it is larger than the a priori criterion. The criterion for deciding what size difference from the mean is sufficiently large as to be not attributed to chance is conventional and usually set at 5% probability.