Robust and nonparametric methods provide an effective way of dealing with the many fundamental problems associated with least squares techniques. Included are valid statistical inferences for situations where models based on normality break down. Nonparametric and robust methods can yield substantial gains in power, highly accurate confidence intervals in situations where classic least squares methods perform poorly, and they provide alternative perspectives that deepen and enhance our understanding of data. Our goal is to bring together the most recent advances in the field. Topics can include, but are not limited to regression, learning models, clustering, ANOVA, outlier detection, techniques for general models, diagnostic methods, and adjustment for covariates.

Co-Chairs:

Rand Wilcox Dept of Psychology University of Southern California Los Angeles, CA 90089-1061, USA E-mail: rwilcox@usc.edu	Simon Sheather Australian Graduate School of Management University of New Souths Wales Sydney, New South Wales Australia E-mail: simonsh@agsm.edu.au
Edgar Brunner Abt. Med. Statistik Universitat Gottingen Humboldt Allee 32 D-37073 Gottingen Germany E-mail: brunner@ams.med.uni-goettingen.de