Paper Summary
Share...

Direct link:

Comparing Bayesian and Frequentist Measurement Invariance Detection Techniques: A Monte Carlo Simulation

Thu, May 4, 8:00 to 9:30am CDT (8:00 to 9:30am CDT), Division D Virtual Sessions, Division D - Section 2: Quantitative Methods and Statistical Theory Virtual Paper Room

Abstract

In the last 10 years there has been an increase in studying measurement invariance. Specifically, Bayesian measurement invariance (MI) techniques, the Bayesian approximate MI (Muthén & Asparouhov, 2013) and Bayesian interval MI (Shi et al., 2018), have been reported to outperform Frequentist approaches. However, the two Bayesian approaches have not been compared in the same simulation study. A Monte Carlo simulation study was conducted varying the sample size per group (n = 50, 100, 200, 300) and magnitude of a non-invariant path coefficient (0, .05, .20, .40). The results from this simulation showed that both Bayesian interval and Frequentist MI performed well under large non-invariance and large sample size conditions, while Bayesian approximate failed to detect non-invariance.

Authors