Comparing RMSEA-Based Indices for Assessing Measurement Invariance in Confirmatory Factor Models

Area of Science:

• Psychometrics
• Statistical Modeling
• Multigroup Confirmatory Factor Analysis (CFA)

Background:

• Fit indices assess how well a confirmatory factor analysis (CFA) model fits data.
• In multigroup models, fit indices evaluate measurement invariance before group comparisons.
• RMSEA_D is an adaptation of RMSEA for assessing nested model consistency across groups.

Purpose of the Study:

• To comprehensively compare the performance of RMSEA_D against ΔRMSEA for evaluating measurement invariance.
• To determine which fit index demonstrates greater sensitivity and detection ability in multigroup CFA.

Main Methods:

• The study involved theoretical derivations of RMSEA_D and ΔRMSEA.
• A population analysis using one-factor CFA models with common research features was conducted.
• Nested models were compared to assess measurement invariance.

Main Results:

• RMSEA_D consistently showed increased sensitivity compared to ΔRMSEA as the number of indicator variables increased.
• RMSEA_D demonstrated a greater ability to detect noninvariance in one-factor models than ΔRMSEA.
• The findings were consistent across both derivations and population analyses.

Conclusions:

• RMSEA_D is recommended over ΔRMSEA for evaluating measurement invariance in multigroup CFA.
• The enhanced sensitivity of RMSEA_D aids in more accurate assessment of invariance.
• This provides researchers with a more reliable tool for cross-group comparisons.