The Impact of Model Size on the Sensitivity of Fit Measures in Measurement Invariance Testing

• In Event: Differential Item Functioning and Measurement Invariance

Abstract

Measurement invariance testing is very important for comparing latent means across groups. Empirical researchers have been using the cut off criteria of fit measures recommended by methodologists to evaluate measurement invariance testing without differentiating model size. However, the model size varies substantially in empirical research, and it is likely that the criteria may not be generalizable to larger model size with multiple factors and multiple items on each factor. This study aims to examine the impact of model size on the sensitivity of fit measures in measurement invariance testing. Preliminary results suggested the change of CFI and RMSEA criterial failed to detect measurement noninvariance when model size was large (e.g., 6 factors with 8 items on each factor).

Authors

• Chunhua Cao, University of Alabama

• Xinya Liang, University of Arkansas