It is shown that the Lack-of-Fit test can be considered as the likelihood ratio test on the mean structure for some linear model. The asymptotic expansion of the null distribution of the test statistic is derived up to order $n^-1$ under nonnormality. A certain robustness against nonnormality is also investigated.
Publié le : 2006-03-14
Classification:
Lack-of-Fit test,
asymptotic expansion,
nonnormality,
multivariate linear model,
null distribution,
62H10,
62H20
@article{1147883394,
author = {Matsumoto, Chieko and Wakaki, Hirofumi},
title = {Asymptotic expansion for Lack-of-Fit test under nonnormality},
journal = {Hiroshima Math. J.},
volume = {36},
number = {1},
year = {2006},
pages = { 29-38},
language = {en},
url = {http://dml.mathdoc.fr/item/1147883394}
}
Matsumoto, Chieko; Wakaki, Hirofumi. Asymptotic expansion for Lack-of-Fit test under nonnormality. Hiroshima Math. J., Tome 36 (2006) no. 1, pp. 29-38. http://gdmltest.u-ga.fr/item/1147883394/