Sankhya: The Indian Journal of Statistics

2002, Volume 64, Series A, Pt. 3, 678--692

RESAMPLING METHODS FOR THE NONPARAMETRIC BEHRENS-FISHER PROBLEM

By

GUTTI JOGESH BABU, Pennsylvania State University, USA
and
A.R. PADMANABHAN, Monash University, Australia

SUMMARY. A robust solution to the non-parametric Behrens-Fisher problem is proposed, without assuming the symmetry of the underlying distributions. The methodology consists of bootstrapping an appropriately centered version of the Mann-Whitney statistic. A theoretical justification of the bootstrap is also presented.

AMS (1991) subject classification}. Primary 62G10; secondary 62G35, 62G09.

Key words and phrases. Bootstrap, Mann-Whitney Statistic.

Full paper (PDF)