A RANK-SUM TEST OF WHETHER TWO MULTIVARIATE SAMPLES WERE DRAWN FROM THE SAME POPULATION.
NEW YORK STATE COLL OF AGRICULTURE ITHACA
Pagination or Media Count:
The problem considered was first posed as the question of whether birds of a feather nest together. The statistical problem is that of devising a test of the hypothesis that the two species were randomly mixed over the occupied cliff ledges. A configuration of Nmn points in a metric space which have been partitioned into two sets, A and B, of m and n points, respectively, is given and a procedure is required for testing the hypothesis of random partitioning against the alternative hypothesis that A and B are the result of some form of segregating operation. Such a nonparametric test procedure, being based only on the conditional distribution of partitions for the given configuration of mn points, should have application to the general problem of testing whether two multivariate samples were drawn from the same population. Author