A Comparison of Four Estimators of a Population Measure of Model Misfit in Covariance Structure Analysis

Master's Thesis
Thumbnail

Abstract

A major issue in the utilization of covariance structure analysis is model misfit evaluation. Recent years have witnessed increasing interest in various test statistics and so-called fit indices, most of which are actually based on or closely related to F0, a measure of model misfit in the population. The present study aims to provide a systematic investigation about the performance of four estimators of F0 available. F01 is the conventional estimator and is based on noncentral chi-square approximation. F02 is newly proposed and does not assume noncentral chi-square approximation. F03 and F04 are derivatives of F02. A Monte Carlo simulation study is conducted to examine how the above four estimators of F0 perform across varying model misspecifications, data distributions, model complexities, and sample sizes. Although all four quantities estimate F0 satisfactorily under normality, the results favor F02 due to its relative robustness to data nonnormality. Issues related to our findings are discussed.

Attributes

Attribute NameValues
URN
  • etd-10272005-175023

Author Wei Zhang
Advisor Ke-Hai Yuan
Contributor Steven Boker, Committee Member
Contributor Ke-Hai Yuan, Committee Chair
Contributor Scott Maxwell, Committee Member
Degree Level Master's Thesis
Degree Discipline Psychology
Degree Name MA
Defense Date
  • 2005-08-31

Submission Date 2005-10-27
Country
  • United States of America

Subject
  • covariance structure analysis

  • model fit

Publisher
  • University of Notre Dame

Language
  • English

Record Visibility and Access Public
Content License
  • All rights reserved

Departments and Units

Files

Please Note: You may encounter a delay before a download begins. Large or infrequently accessed files can take several minutes to retrieve from our archival storage system.