A test statistic used when we are interested in the null hypothesis that two or more samples were drawn from populations with identical means.