A Bayesian approach to nonlinear latent variable models using the Gibbs sampler and the Metropolis-Hastings algorithm

被引:153
作者
Arminger, G
机构
[1] Berg Univ Gesamthsch Wuppertal, Dept Econ, D-42097 Wuppertal, Germany
[2] Univ Calif Los Angeles, Grad Sch Educ & Informat Studies, Los Angeles, CA 90024 USA
关键词
Gibbs sampler; LISREL model; Metropolis-Hastings algorithm; non-linear functions; of latent regressors;
D O I
10.1007/BF02294856
中图分类号
O1 [数学];
学科分类号
0701 ; 070101 ;
摘要
Nonlinear latent variable models are specified that include quadratic forms and interactions of latent regressor variables as special cases. To estimate the parameters, the models are put in a Bayesian framework with conjugate priors for the parameters. The posterior distributions of the parameters and the latent variables are estimated using Markov chain Monte Carlo methods such as the Gibbs sampler and the Metropolis-Hastings algorithm. The proposed estimation methods are illustrated by two simulation studies and by the estimation of a non-linear model for the dependence of performance on task complexity and goal specificity using empirical data.
引用
收藏
页码:271 / 300
页数:30
相关论文
共 37 条
[1]  
AHYDUK LA, 1987, STRUCTURAL EQUATION
[2]  
Anderson T, 1984, INTRO MULTIVARIATE
[3]  
[Anonymous], 1995, Markov Chain Monte Carlo in Practice
[4]   PSEUDO MAXIMUM-LIKELIHOOD ESTIMATION AND A TEST FOR MISSPECIFICATION IN MEAN AND COVARIANCE STRUCTURE MODELS [J].
ARMINGER, G ;
SCHOENBERG, RJ .
PSYCHOMETRIKA, 1989, 54 (03) :409-425
[5]  
ARMINGER G, 1996, MECOSA3 USER GUIDE
[6]  
ARNOLD SF, 1993, HDB STAT, V9, P599
[7]  
BESAG J, 1995, STAT SCI, V19, P3
[8]  
Box GE., 2011, BAYESIAN INFERENCE S
[10]  
CALRIN BP, 1996, BAYES EMPIRICAL BAYE