Premium
Inference about covariances under missing values
Author(s) -
Provost Serge B.
Publication year - 1991
Publication title -
environmetrics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.68
H-Index - 58
eISSN - 1099-095X
pISSN - 1180-4009
DOI - 10.1002/env.3770020407
Subject(s) - mathematics , statistics , test statistic , independence (probability theory) , asymptotic distribution , statistic , sample size determination , likelihood ratio test , missing data , uncorrelated , statistical hypothesis testing , inference , variable (mathematics) , basis (linear algebra) , estimator , mathematical analysis , computer science , artificial intelligence , geometry
The likelihood ratio criterion for testing the mutual independence of q subvectors of an l ‐dimentional normal vector on the basis of a sample of size N on l components and of an incomplete sample of size M on r components ( r < l ). This is equivalent to testing the hypothesis that each variable in one subvector is uncorrelated with each variable in the other subvectors. This asymptotic distribution of the test statistic is given.