A presentation of some nonparametric tests of whether the [r] largest observations of the set are too large or too small to be consistent with the hypothesis that these populations have a common median value. The observations are statistically indepe...