论文信息 - Stabilized multivariate tests : the inclusion of missing values

Stabilized multivariate tests : the inclusion of missing values

In two recent papers, LAUTER (1996) and LAUTER, GLIMM, and KROPF (1996) have proposed a new class of exact tests for multivariate normal data with an inherent structure. These tests are based on the construction of scores that have a spherical distribution. The principle can be adapted to a variety of different situations, and it supports exact tests with small samples of high-dimensional data. In this paper, an enhancement of that theory will be given regarding the problem of missing data. It provides a new mathematical understanding of mean imputation techniques, and it is also the basis for a modified principal component missing data algorithm described here.

Jürgen Läuter | Siegfried Kropf | E. Glimm

[1] Roderick J. A. Little,et al. Statistical Analysis with Missing Data , 1988 .

[2] A. L. Bello. A Bootstrap Method for Using Imputation Techniques for Data With Missing Values , 1994 .

[3] Jürgen Läuter. Stabile multivariate Verfahren : Diskriminanzanalyse - Regressionsanalyse - Faktoranalyse , 1992 .

[4] P. O'Brien. Procedures for comparing samples with multiple endpoints. , 1984, Biometrics.

[5] Siegfried Kropf,et al. New Tests for Data with an Inherent Structure , 1996 .

[6] R. Fisher. THE SAMPLING DISTRIBUTION OF SOME STATISTICS OBTAINED FROM NON‐LINEAR EQUATIONS , 1939 .

[7] Jürgen Läuter,et al. New multivariate tests for data with an inherent structure , 1996 .

[8] K. Fang,et al. Generalized Multivariate Analysis , 1990 .

[9] Jurgen Lauter,et al. Exact t and F Tests for Analyzing Studies with Multiple Endpoints , 1996 .