Testing Measurement Invariance and Factor Mean Difference With Many Groups: Empirical Comparison of Five Approaches

• In Event: Poster Session 11
  In Poster Session: Common Latent Variable Modeling Issues and Potential Answers

Abstract

With the increasing use of international survey data especially in cross-cultural and multi-national studies, establishing measurement invariance (MI) across a large number of groups in a study is essential. However, testing MI over many groups is methodologically challenging. We introduce five methods for MI testing across many groups: multiple group confirmatory factor analysis, multilevel confirmatory factor analysis, multilevel factor mixture model, Bayesian approximate MI (Muthén & Asparouhov, 2013), and alignment (Asparouhov & Muthén, 2014). The procedures of MI testing and subsequent multiple group factor mean comparisons with these five methods are demonstrated with empirical data. A Monte Carlo study is conducted to investigate the performance of these methods. Practical implications in applied research are discussed.

Authors

• Eun Sook Kim, University of South Florida

• Chunhua Cao, Polk County School District

• Yan Wang, University of South Florida

• Diep Thi Nguyen, University of South Florida