Principal component analysis of binary data by iterated singular value decomposition

被引：61

作者：

de Leeuw, J ^{[1
]}

机构：

[1] Univ Calif Los Angeles, Dept Stat, Los Angeles, CA 90095 USA

来源：

COMPUTATIONAL STATISTICS & DATA ANALYSIS | 2006年 / 50卷 / 01期

关键词：

multivariate analysis; factor analysis; binary data; item response models; applications to social sciences;

D O I：

10.1016/j.csda.2004.07.010

中图分类号：

TP39 [计算机的应用];

学科分类号：

081203 ; 0835 ;

摘要：

The maximum-likelihood estimates of a principal component analysis on the logit or probit scale are computed using majorization algorithms that iterate a sequence of weighted or unweighted singular value decompositions. The relation with similar methods in item response theory, roll call analysis, and binary choice analysis is discussed. The technique is applied to 2001 US House roll call data. (c) 2004 Elsevier B.V. All rights reserved.

引用

页码：21 / 39

页数：19

共 47 条

[1]

*ADA, 2002, ADA TODAY, V57, P1

[2]

ANDERSON TW, 1956, 3RD P BERK S MATH ST, V5, P111

[3]

[Anonymous], 1955, Engineering Research Bulletin

[4]

[Anonymous], 1966, MULTIVARIATE ANAL P

[5]

[Anonymous], 1994, INFORM SYSTEMS DATA

[6]

[Anonymous], 1990, NONLINEAR MULTIVARIA

[7]

Bailey D M, 2000, High Alt Med Biol, V1, P9, DOI 10.1089/152702900320649

[8]

Bartholomew DJ., 1987, LATENT VARIABLE MODE

[9] MCMC estimation and some model-fit analysis of multidimensional IRT models [J].

Béguin, AA ;

Glas, CAW .

PSYCHOMETRIKA, 2001, 66 (04) :541-561

[10] FULL-INFORMATION ITEM FACTOR-ANALYSIS [J].

BOCK, RD ;

GIBBONS, R ;

MURAKI, E .

APPLIED PSYCHOLOGICAL MEASUREMENT, 1988, 12 (03) :261-280

← 1 2 3 4 5 →