Evaluating estimation methods for ordinal data in structural equation modeling

Research output: Contribution to journalArticle

83 Citations (Scopus)

Abstract

This study examined the performance of two alternative estimation approaches in structural equation modeling for ordinal data under different levels of model misspecification, score skewness, sample size, and model size. Both approaches involve analyzing a polychoric correlation matrix as well as adjusting standard error estimates and model chi-squared, but one estimates model parameters with maximum likelihood and the other with robust weighted least-squared. Relative bias in parameter estimates and standard error estimates, Type I error rate, and empirical power of the model test, where appropriate, were evaluated through Monte Carlo simulations. These alternative approaches generally provided unbiased parameter estimates when the model was correctly specified. They also provided unbiased standard error estimates and adequate Type I error control in general unless sample size was small and the measured variables were moderately skewed. Differences between the methods in convergence problems and the evaluation criteria, especially under small sample and skewed variable conditions, were discussed.

Original languageEnglish (US)
Pages (from-to)495-507
Number of pages13
JournalQuality and Quantity
Volume43
Issue number3
DOIs
StatePublished - May 1 2009

Fingerprint

Ordinal Data
Structural Equation Modeling
Standard error
Error Estimates
Sample Size
Polychoric Correlation
Estimate
Chi-squared
Model Misspecification
Type I Error Rate
Error Model
Type I error
Alternatives
Error Control
Correlation Matrix
Skewness
Small Sample
Model
Maximum Likelihood
Standard Model

All Science Journal Classification (ASJC) codes

  • Statistics and Probability
  • Social Sciences(all)

Cite this

@article{d1c7342b2029436f9e647116f15a4e16,
title = "Evaluating estimation methods for ordinal data in structural equation modeling",
abstract = "This study examined the performance of two alternative estimation approaches in structural equation modeling for ordinal data under different levels of model misspecification, score skewness, sample size, and model size. Both approaches involve analyzing a polychoric correlation matrix as well as adjusting standard error estimates and model chi-squared, but one estimates model parameters with maximum likelihood and the other with robust weighted least-squared. Relative bias in parameter estimates and standard error estimates, Type I error rate, and empirical power of the model test, where appropriate, were evaluated through Monte Carlo simulations. These alternative approaches generally provided unbiased parameter estimates when the model was correctly specified. They also provided unbiased standard error estimates and adequate Type I error control in general unless sample size was small and the measured variables were moderately skewed. Differences between the methods in convergence problems and the evaluation criteria, especially under small sample and skewed variable conditions, were discussed.",
author = "Pui-wa Lei",
year = "2009",
month = "5",
day = "1",
doi = "10.1007/s11135-007-9133-z",
language = "English (US)",
volume = "43",
pages = "495--507",
journal = "Quality and Quantity",
issn = "0033-5177",
publisher = "Springer Netherlands",
number = "3",

}

Evaluating estimation methods for ordinal data in structural equation modeling. / Lei, Pui-wa.

In: Quality and Quantity, Vol. 43, No. 3, 01.05.2009, p. 495-507.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Evaluating estimation methods for ordinal data in structural equation modeling

AU - Lei, Pui-wa

PY - 2009/5/1

Y1 - 2009/5/1

N2 - This study examined the performance of two alternative estimation approaches in structural equation modeling for ordinal data under different levels of model misspecification, score skewness, sample size, and model size. Both approaches involve analyzing a polychoric correlation matrix as well as adjusting standard error estimates and model chi-squared, but one estimates model parameters with maximum likelihood and the other with robust weighted least-squared. Relative bias in parameter estimates and standard error estimates, Type I error rate, and empirical power of the model test, where appropriate, were evaluated through Monte Carlo simulations. These alternative approaches generally provided unbiased parameter estimates when the model was correctly specified. They also provided unbiased standard error estimates and adequate Type I error control in general unless sample size was small and the measured variables were moderately skewed. Differences between the methods in convergence problems and the evaluation criteria, especially under small sample and skewed variable conditions, were discussed.

AB - This study examined the performance of two alternative estimation approaches in structural equation modeling for ordinal data under different levels of model misspecification, score skewness, sample size, and model size. Both approaches involve analyzing a polychoric correlation matrix as well as adjusting standard error estimates and model chi-squared, but one estimates model parameters with maximum likelihood and the other with robust weighted least-squared. Relative bias in parameter estimates and standard error estimates, Type I error rate, and empirical power of the model test, where appropriate, were evaluated through Monte Carlo simulations. These alternative approaches generally provided unbiased parameter estimates when the model was correctly specified. They also provided unbiased standard error estimates and adequate Type I error control in general unless sample size was small and the measured variables were moderately skewed. Differences between the methods in convergence problems and the evaluation criteria, especially under small sample and skewed variable conditions, were discussed.

UR - http://www.scopus.com/inward/record.url?scp=64749097255&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=64749097255&partnerID=8YFLogxK

U2 - 10.1007/s11135-007-9133-z

DO - 10.1007/s11135-007-9133-z

M3 - Article

AN - SCOPUS:64749097255

VL - 43

SP - 495

EP - 507

JO - Quality and Quantity

JF - Quality and Quantity

SN - 0033-5177

IS - 3

ER -