Analysis of multisample identified and non-identified structural equation models with stochastic constraints

Sik Yum Lee, Wai Tung Ho

Research output: Contribution to journalArticlepeer-review

5 Citations (Scopus)

Abstract

In this paper, multisample structural equation models with stochastic constrains are investigated using a Bayesian approach. Both identified and non-identified models are considered. Asymptotic properties of the Bayesian estimates are developed and a scoring type algorithm is implemented to obtain the solution. A simulation study is conducted to examine the empirical behaviors of the Bayesian estimates. The conclusion is that it is advantageous to incorporate stochastic constraints in the analysis of the model.

Original languageEnglish
Pages (from-to)441-453
Number of pages13
JournalComputational Statistics and Data Analysis
Volume16
Issue number4
DOIs
Publication statusPublished - Oct 1993

Keywords

  • Asymptotic properties
  • Bayesian estimates
  • Scoring algorithm
  • Simulation study

Fingerprint

Dive into the research topics of 'Analysis of multisample identified and non-identified structural equation models with stochastic constraints'. Together they form a unique fingerprint.

Cite this