Simple robust tests for scale differences in paired data

SUMMARY The classical statistical test for assessing the difference between the two scale parameters in paired data due to Pitman (1939) and Morgan (1939) is not robust. This paper explores robust alternatives to Pitman's test, using the framework of the one-sample t-test. The asymptotic behaviour of the tests under the null hypothesis is examined. In general, it is easy to construct tests that are asymptotically distribution-free, provided the data come from a symmetric bivariate distribution. Versions of these tests that do not require symmetry are derived. Monte Carlo simulation experiments are done to assess small sample behaviour and power characteristics. The tests are applied to a small data set comparing cancerous and noncancerous lungs.

[1]  M. Kendall.  Rank Correlation Methods , 1949 .

[2]  Alan J. Lee,et al.  U-Statistics: Theory and Practice , 1990 .

[3]  E. Pitman.  A NOTE ON NORMAL CORRELATION , 1939 .

[4]  M. E. Johnson,et al.  A Family of Distributions for Modelling Non‐Elliptically Symmetric Multivariate Data , 1981 .

[5]  Vic Barnett,et al.  Some bivariate uniform distributions , 1980 .

[6]  R. Randles,et al.  Introduction to the Theory of Nonparametric Statistics , 1991 .

[7]  M. Kendall,et al.  Kendall's Advanced Theory of Statistics: Volume 1 Distribution Theory , 1987 .

[8]  On the Use of Conditional Asymptotic Normality , 1979 .

[9]  P. Harris.  Testing for variance homogeneity of correlated variables , 1985 .

[10]  R. Douglas Martin,et al.  S-PLUS Version 3 , 1992 .

[11]  G. J. Babu,et al.  A Note on Bootstrapping the Sample Median , 1984 .

[12]  Charles E. McCulloch,et al.  Tests for Equality of Variances with Paired Data , 1987 .

[13]  M. E. Johnson,et al.  Generalized Burr-Pareto-Logistic distributions with applications to a uranium exploration data set , 1986 .

[14]  J. Shao,et al.  A General Theory for Jackknife Variance Estimation , 1989 .

[15]  Ronald H. Randles,et al.  On the Asymptotic Normality of Statistics with Estimated Parameters , 1982 .

[16]  W. Hoeffding.  A Class of Statistics with Asymptotically Normal Distribution , 1948 .

[17]  W. A. Morgan.  TEST FOR THE SIGNIFICANCE OF THE DIFFERENCE BETWEEN THE TWO VARIANCES IN A SAMPLE FROM A NORMAL BIVARIATE POPULATION , 1939 .

[18]  P. Grambsch,et al.  A prospective pilot study evaluating the effectiveness of secretory IgA measurements in bronchoalveolar lavage to detect non-small cell lung cancer. , 1990, Chest.

[19]  W. R. Buckland,et al.  Contributions to Probability and Statistics , 1960 .

[20]  Morton B. Brown,et al.  Robust Tests for the Equality of Variances , 1974 .

[21]  C. Genest,et al.  The Joy of Copulas: Bivariate Distributions with Uniform Marginals , 1986 .

[22]  R. Iman,et al.  Rank Transformations as a Bridge between Parametric and Nonparametric Statistics , 1981 .

[23]  B. Olsson,et al.  A nearly distribution-free test for comparing dispersion in paired samples , 1982 .

[24]  R. Wilcox.  Comparing the variances of dependent groups , 1989 .