The author presents a multivariate location model for cluster correlated observations. He proposes an affine-invariant multivariate sign statistic for testing the value of the location parameter. His statistic is an adaptation of that proposed by Randles (2000). The author shows, under very mild conditions, that his test statistic is asymptotically distributed as a chi-squared random variable under the null hypothesis. In particular, the test can be used for skewed populations. In the context of a general multivariate normal model, the author obtains values of his test's Pitman asymptotic efficiency relative to another test based on the overall average. He shows that there is an improvement in the relative performance of the new test as soon as intra-cluster correlation is present Even in the univariate case, the new test can be very competitive for Gaussian data. Furthermore, the statistic is easy to compute, even for large dimensional data. The author shows through simulations that his test performs well compared to the average-based test. He illustrates its use with real data.
L'auteur presente un modele de position multivarie pour donnees correlees en grappes. Il propose une statistique du signe multivarie affine-invariant permettant de tester la valeur du vecteur de position. Sa statistique est une adaptation de celle proposee par Randles (2000). L'auteur montre que sous des conditions peu restrictives, la loi asymptotique de sa statistique sous l'hypothese nulle est celle du khi-deux. En particulier, le test peut ětre utilise avec des populations asymetriques. Dans le cadre d'un modele multinormal general, l'auteur calcule les valeurs de l'efficacite asymptotique de Pitman de son test par rapport a un autre test base sur la moyenne globale. Ses resultats montrent que la performance du nouveau test s'ameliore en presence de correlation intra-grappe. Měme dans le cas univarie, le nouveau test s'avere tres performant pour des donnees gaussiennes. De plus, la statistique se calcule facilement, měme en haute dimension. L'auteur montre par simulation que son test se comporte bien par rapport a celui fonde sur la moyenne globale. Il en illustre l'emploi au moyen de donnees reelles.
[1]
R. Randles,et al.
A practical affine equivariant multivariate median
,
2002
.
[2]
R. Randles.
A Distribution-Free Multivariate Sign Test Based on Interdirections
,
1989
.
[3]
J. G. Saw,et al.
Dependent unit vectors
,
1983
.
[4]
P. Sen,et al.
Chapter 3 – Elementary theory of rank tests
,
1999
.
[5]
Yasuo Amemiya,et al.
Asymptotic Properties of the Estimators for Multivariate Components of Variance
,
1994
.
[6]
B. M. Brown,et al.
An Affine Invariant Bivariate Version of the Sign Test
,
1989
.
[7]
Ronald H. Randles,et al.
A Simpler, Affine-Invariant, Multivariate, Distribution-Free Sign Test
,
2000
.
[8]
R. Tremblay,et al.
DSM-III-R and DSM-III criteria for conduct disorder in preadolescent girls: specific but insensitive.
,
1996,
Journal of the American Academy of Child and Adolescent Psychiatry.
[9]
Hannu Oja,et al.
Multivariate spatial sign and rank methods
,
1995
.
[10]
André I. Khuri,et al.
Statistical Texts for Mixed Linear Models: Khuri/Statistical
,
1998
.
[11]
David E. Tyler.
A Distribution-Free $M$-Estimator of Multivariate Scatter
,
1987
.
[12]
J. L. Hodges,et al.
A Bivariate Sign Test
,
1955
.
[13]
Denis Larocque,et al.
Bivariate Sign Tests Based on the Sup, L1 and L2 Norms
,
2000
.
[14]
Hannu Oja,et al.
ON THE EFFICIENCY OF MULTIVARIATE SPATIAL SIGN AND RANK TESTS
,
1997
.
[15]
B. M. Brown,et al.
Statistical Uses of the Spatial Median
,
1983
.
[16]
E. JacquelinDietz,et al.
Linear signed rank tests for hultivariate location
,
1984
.
[17]
Hannu Oja,et al.
Affine Invariant Multivariate One‐Sample Sign Tests
,
1994
.
[18]
Hannu Oja,et al.
Bivariate Sign Tests
,
1989
.
[19]
R. Tremblay,et al.
Behavioural development in children of divorce and remarriage.
,
1997,
Journal of child psychology and psychiatry, and allied disciplines.
[20]
I. Blumen,et al.
A New Bivariate Sign Test
,
1958
.
[21]
Hannu Oja,et al.
OPERATING TRANSFORMATION RETRANSFORMATION ON SPATIAL MEDIAN AND ANGLE TEST
,
1998
.