Article
Keywords:
asymptotic normality of multivariate linear rank statistics; general alternatives
Summary:
Let $X_j, 1\leq j\leq N$, be independent random $p$-vectors with respective continuous cumulative distribution functions $F_j 1\leq j\leq N$. Define the $p$-vectors $R_j$ by setting $R_{ij}$ equal to the rank of $X_{ij}$ among $X_{ij}, \ldots, X_{iN}, 1\leq i \leq p, 1\leq j \leq N$. Let $a^{(N)}(.)$ denote a multivariate score function in $R_p$, and put $S= \sum ^N_{j=1} c_ja^{(N)}(R_j)$, the $c_j$ being arbitrary regression constants.
In this paper the asymptotic distribution of $S$ is investigated under various sets of conditions on the constants, the score functions, and the underlying distribution functions. In particular, asymptotic normality of $S$ is established under the circumstance that the $F_j$ are merely continuous. In addition, under mild conditions, centering vectors for $S$ are found.
References:
[1] Chernoff H., and Savage I. R.:
Asymptotic normality and efficiency of certain nonparametric test statistics. Ann. Math. Stat. 29 (1958), 972-994.
DOI 10.1214/aoms/1177706436 |
MR 0100322
[2] Dupač V.:
A contribution to the asymptotic normality of simple linear rank statistics. In Nonparametric Techniques in Statistical Inference (M. L. Prui, Ed.), pp. 75-88, University Press, Cambridge, 1970.
MR 0283930
[5] Natanson I. P.:
Theory of Functions of a Real Variable 1. Frederick Ungar, New York, 1961.
MR 0067952
[7] Puri M. L., Sen P. K.:
Nonparametric Methods in Multivariate Analysis. John Wiley, New York, 1971.
MR 0298844 |
Zbl 0237.62033
[8] Sen P. K., Puri M. L.:
On the theory of rank order tests for location in the multivariate one sample problem. Ann. Math. Stat. 38 (1968), 1216-1228.
DOI 10.1214/aoms/1177698790 |
MR 0212954