The Impact of Measurement Noninvariance on Latent Change Score Modeling: A Monte Carlo Simulation Study

• In Event: Advancements in Detecting Differential Item Functioning and Measurement Invariance

Abstract

Latent score change modeling (LCSM) combines latent growth model and time-series regression model (McArdle & Nesselroade, 2014). Measurement invariance (MI) over time is critical in the estimation of LCSM. In this simulation study we investigate the impact of measurement noninvariance (MNI) on the estimation of LCSM manipulating sample size and the magnitude, direction, location, and proportion of MNI when proportional change model and dual change model are used with mean composites or MI-assumed measurement models. The results show that even noninvariance in one item could yield severe bias in the LCSM parameter estimates. However, the impact depends on the simulation factors and the parameters (e.g., proportional or constant change). Discussions and implications of findings are presented.

Authors

• Eunsook Kim, University of South Florida

• Yan Wang, University of Massachusetts Lowell

• Siyu Liu, University of South Florida - Tampa