Home > Conventions > Program Archive > 31st APS Annual Convention > Presentation
APS

31st APS Annual Convention · 2019

How Missing Data Affect Fit Indices in Structural Equation Modeling

Washington, DC · May 2019

Poster · Methodology

  • Xijuan Zhang
    University of British Columbia
  • Victoria Savalei
    University of British Columbia

Abstract

Full-information maximum likelihood (FIML) is one of the most common techniques for handling missing data in structural equation modelling. In this poster, we show that fit indices (FIs) can be distorted under FIML estimation. We provide mathematical derivations, analytical examples and simulation studies to demonstrate how missing data affect FIs.

Statistics and Methodology

← Poster Session XIV