MAXIMUM-LIKELIHOOD VARIANCE-COMPONENTS ESTIMATION FOR BINARY DATA

被引：157

作者：

MCCULLOCH, CE ^{[1
]}

机构：

[1] CORNELL UNIV,CTR STAT,ITHACA,NY 14853

来源：

JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION | 1994年 / 89卷 / 425期

关键词：

EM ALGORITHM; FIXED AND RANDOM EFFECTS; GENERALIZED LINEAR MODELS; MONTE-CARLO MARKOV CHAIN; PROBIT;

D O I：

10.2307/2291229

中图分类号：

O21 [概率论与数理统计]; C8 [统计学];

学科分类号：

020208 ; 070103 ; 0714 ;

摘要：

We consider a class of probit normal models for binary data and describe ML and REML estimation of variance components for that class as well as best prediction for the realized values of the random effects. ML estimates are calculated using an EM algorithm; for complicated models EM includes a Gibbs step. The computations are illustrated through two examples.

引用

页码：330 / 335

页数：6

共 40 条

[1]

Abramowitz M., 1964, HDB MATH FUNCTIONS

[2]

ANDERSON DA, 1985, J R STAT SOC B, V47, P203

[3]

Casella G, 1992, VARIANCE COMPONENTS

[4]

Crowder M. J., 1978, J R STAT SOC C-APPL, V27, P34, DOI [DOI 10.2307/2346223, 10.2307/2346223]