|
Organizers |
A Robust Bootstrap Test for the Equality of Several Medians
by
Angelo J. Canty
McMaster University, Hamilton
Coauthors: A. R. Padmanabhan and G. J. Babu
Babu and Padmanabhan (2002) proposed a robust bootstrap solution to the Behrens-Fisher problem. We show that this method can be extended to the test of equality of several medians. The basic assumption is that the data come from a location-scale family but no assumptions on the shape of the underlying distribution are made. An extensive Monte Carlo simulation study will be presented to illustrate the power of the test against umbrella alternatives (Mack and Wolfe, 1981) and pattern alternatives (Hettmansperger and Norton, 1987). We shall also give some theoretical results to justify the method asymptotically. These results constitute a multi-sample generalization of those in Babu and Padmanabhan (2002) and the asymptotics follow along the same lines as in that paper.
Date received: September 17, 2003
Copyright © 2003 by the author(s). The author(s) of this document and the organizers of the conference have granted their consent to include this abstract in Atlas Mathematical Conference Abstracts. Document # came-73.