Small sample inference for fixed effects from restricted maximum likelihood

被引:3463
作者
Kenward, MG [1 ]
Roger, JH [1 ]
机构
[1] LIVE DATA PROC,CHINNOR OX9 4BQ,OXON,ENGLAND
关键词
alpha design; ante-dependence; crossover trial; mixed models; residual maximum likelihood; small sample approximation;
D O I
10.2307/2533558
中图分类号
Q [生物科学];
学科分类号
07 ; 0710 ; 09 ;
摘要
Restricted maximum likelihood (REML) is now well established as a method for estimating the parameters of the general Gaussian linear model with a structured covariance matrix, in particular for mixed linear models. Conventionally, estimates of precision and inference for fixed effects are based on their asymptotic distribution, which is known to be inadequate for some small-sample problems. In this paper, we present a scaled Wald statistic, together with an F approximation to its sampling distribution, that is shown to perform well in a range of small sample settings. The statistic uses an adjusted estimator of the covariance matrix that has reduced small sample bias. This approach has the advantage that it reproduces both the statistics and F distributions in those settings where the latter is exact, namely for Hotelling T-2 type statistics and for analysis of variance F-ratios. The performance of the modified statistics is assessed through simulation studies of four different REML analyses and the methods are illustrated using three examples.
引用
收藏
页码:983 / 997
页数:15
相关论文
共 26 条
[21]  
McLean R.A., 1988, P STAT COMPUTING SEC, P50
[22]   RECOVERY OF INTER-BLOCK INFORMATION WHEN BLOCK SIZES ARE UNEQUAL [J].
PATTERSON, HD ;
THOMPSON, R .
BIOMETRIKA, 1971, 58 (03) :545-+
[23]  
Robinson G.K., 1991, Statistical science, V6, P15, DOI DOI 10.1214/SS/1177011926
[24]   SYNTHESIS OF VARIANCE [J].
Satterthwaite, Franklin E. .
PSYCHOMETRIKA, 1941, 6 (05) :309-316
[25]  
Thompson R., 1980, Mathematische Operationsforschung und Statistik, Series Statistics, V11, P545, DOI 10.1080/02331888008801560
[26]  
WELHAM SJ, 1997, IN PRESS J ROYAL S B