Joint asymptotic multinormality for a class of rank order statistics in multivariate paired comparisons
The joint asymptotic multinormality of certain linear signed-rank statistics introduced by Shane and Puri (1969) is established for the nonidentically distributed case; moreover, the usual restriction forbidding constant score generating functions is dropped. In addition, sufficient conditions more general than those of Shane and Puri are given for the convergence of certain dispersion matrices, and these conditions guarantee the asymptotic independence of the statistics under consideration.
Year of publication: |
1974
|
---|---|
Authors: | Russell, Carl T. ; Puri, Madan L. |
Published in: |
Journal of Multivariate Analysis. - Elsevier, ISSN 0047-259X. - Vol. 4.1974, 1, p. 88-105
|
Publisher: |
Elsevier |
Keywords: | Signed rank statistics joint asymptotic multinormality Dispersion matrix asymptotic independence |
Saved in:
Saved in favorites
Similar items by person
-
Some aspects of nonparametric inference
Puri, Madan L., (1972)
-
Testing hypotheses about the equality of several risk measure values with applications in insurance
Jones, Bruce L., (2006)
-
Weak invariance of generalized u-statistics for nonstationary absolutely regular processes
Harel, Michel, (1990)
- More ...