Nonparametric multivariate rank tests and their unbiasedness

Although unbiasedness is a basic property of a good test, many tests on vector parameters or scalar parameters against two-sided alternatives are not finite-sample unbiased. This was already noticed by Sugiura [Ann. Inst. Statist. Math. 17 (1965) 261--263]; he found an alternative against which the Wilcoxon test is not unbiased. The problem is even more serious in multivariate models. When testing the hypothesis against an alternative which fits well with the experiment, it should be verified whether the power of the test under this alternative cannot be smaller than the significance level. Surprisingly, this serious problem is not frequently considered in the literature. The present paper considers the two-sample multivariate testing problem. We construct several rank tests which are finite-sample unbiased against a broad class of location/scale alternatives and are finite-sample distribution-free under the hypothesis and alternatives. Each of them is locally most powerful against a specific alternative of the Lehmann type. Their powers against some alternatives are numerically compared with each other and with other rank and classical tests. The question of affine invariance of two-sample multivariate tests is also discussed.

[1]  J. Oosterhoff,et al.  A Note on Contiguity and Hellinger Distance , 2012 .

[2]  Biased and Unbiased Two-sided Wilcoxon Tests for Equal Sample Sizes , 2006 .

[3]  N. Henze,et al.  On the multivariate runs test , 1999 .

[4]  L1-Derivatives, Score Functions and Tests , 2002 .

[5]  P. Sen,et al.  Nonparametric methods in multivariate analysis , 1974 .

[6]  Georg Ch. Pflug,et al.  Mathematical statistics and applications , 1985 .

[7]  L. Baringhaus,et al.  On a new multivariate two-sample test , 2004 .

[8]  The locally unbiased two-sided Durbin—Watson test , 1991 .

[9]  J. Friedman,et al.  Multivariate generalizations of the Wald--Wolfowitz and Smirnov two-sample tests , 1979 .

[10]  M. Schilling.  Multivariate Two-Sample Tests Based on Nearest Neighbors , 1986 .

[11]  J. E. Walsh,et al.  Contributions to the Theory of Rank Order Statistics--The Two Sample Case , 1958 .

[12]  Hannu Oja,et al.  ON THE EFFICIENCY OF MULTIVARIATE SPATIAL SIGN AND RANK TESTS , 1997 .

[13]  Robert L. Obenchain.  Multivariate Procedures Invariant Under Linear Transformations , 1971 .

[14]  B. M. Brown,et al.  Cramer-von Mises distributions and permutation tests , 1982 .

[15]  Erich L. Lehmann,et al.  The Power of Rank Tests , 1953 .

[16]  Inference based on the affine invariant multivariate Mann–Whitney–Wilcoxon statistic , 2003 .

[17]  An example of a two-sided Wilcoxon signed rank test which is not unbiased , 1995 .

[18]  P. Bickel.  A Distribution Free Version of the Smirnov Two Sample Test in the $p$-Variate Case , 1969 .

[19]  E. Lehmann.  Testing Statistical Hypotheses , 1960 .

[20]  H. Oja.  Multivariate Nonparametric Methods with R: An approach based on spatial signs and ranks , 2010 .

[21]  H. Oja.  Multivariate Nonparametric Methods with R , 2010 .

[22]  J. Marden,et al.  Test of multivariate linear models using spatial concordances , 2005 .

[23]  R. Y. Liu,et al.  On a notion of simplicial depth. , 1988, Proceedings of the National Academy of Sciences of the United States of America.

[24]  Jaroslav Hájek,et al.  Theory of rank tests , 1969 .

[25]  Stephen E. Fienberg,et al.  Testing Statistical Hypotheses , 2005 .

[26]  J. Marden,et al.  An Approach to Multivariate Rank Tests in Multivariate Analysis of Variance , 1997 .

[27]  Xuming He,et al.  On the limiting distributions of multivariate depth-based rank sum statistics and related tests , 2006, 0708.0167.

[28]  Regina Y. Liu.  On a Notion of Data Depth Based on Random Simplices , 1990 .

[29]  Jon A. Wellner,et al.  Permutation Tests for Directional Data , 1979 .

[30]  P. Sen,et al.  Theory of rank tests , 1969 .

[31]  Regina Y. Liu,et al.  A Quality Index Based on Data Depth and Multivariate Rank Tests , 1993 .

[32]  N. Henze.  A MULTIVARIATE TWO-SAMPLE TEST BASED ON THE NUMBER OF NEAREST NEIGHBOR TYPE COINCIDENCES , 1988 .

[33]  Hannu Oja.  On Permutation Tests in Multiple Regression and Analysis of Covariance Problems , 1987 .

[34]  P. Chaudhuri,et al.  Sign Tests in Multidimension: Inference Based on the Geometry of the Data Cloud , 1993 .

[35]  J. Gibbons.  A Proposed Two‐Sample Rank Test: The Psi Test and its Properties , 1964 .

[36]  Hannu Oja,et al.  Multivariate Nonparametric Tests , 2004 .

[37]  An example of the two-sided wilcoxon test which is not unbiased , 1965 .

[38]  P. Hall,et al.  Permutation tests for equality of distributions in high‐dimensional settings , 2002 .

[39]  J. Tukey.  Mathematics and the Picturing of Data , 1975 .

[40]  D. Paindaveine,et al.  Optimal tests for multivariate location based on interdirections and pseudo-Mahalanobis ranks , 2002 .

[41]  R. Bartoszynski,et al.  Reducing multidimensional two-sample data to one-dimensional interpoint comparisons , 1996 .

[42]  P. Rosenbaum.  An exact distribution‐free test comparing two multivariate distributions based on adjacency , 2005 .

[43]  Ronald H. Randles,et al.  Multivariate rank tests for the two-sample location problem , 1990 .

[44]  Derivative in the mean of a density and statistical applications , 2003 .

[45]  Lixing Zhu,et al.  Permutation Tests for Multivariate Location Problems , 1999 .

[46]  Hannu Oja,et al.  AFFINE INVARIANT MULTIVARIATE RANK TESTS FOR SEVERAL SAMPLES , 1998 .