Dynamic measurement invariance cutoffs for two-group fit index differences.

Area of Science:

• Psychometrics
• Statistical Modeling
• Cross-cultural Research

Background:

• Measurement invariance ensures scales function similarly across groups.
• Current methods use fixed cutoffs for fit index differences (e.g., ΔCFI, ΔRMSEA).
• These fixed cutoffs often fail to generalize due to model and data variability.

Purpose of the Study:

• To address limitations of fixed cutoffs in measurement invariance testing.
• To propose and validate a novel method for deriving dynamic measurement invariance (DMI) cutoffs.
• To enhance the accuracy and generalizability of cross-group measurement comparisons.

Main Methods:

• Review of methodological literature on measurement invariance.
• Extension of dynamic fit index cutoffs to multi-group models.
• Simulation studies using researcher-specific model and data characteristics.
• Development of open-source software for DMI cutoff calculation.

Main Results:

• Fixed cutoffs for ΔCFI and ΔRMSEA demonstrate poor generalizability.
• DMI cutoffs are customized based on specific model and data properties.
• Simulations and empirical examples show DMI's potential for improved invariance assessment.
• Provided software enhances the method's accessibility and utility.

Conclusions:

• Dynamic measurement invariance (DMI) cutoffs offer a more robust approach than fixed thresholds.
• Tailoring cutoffs to specific study parameters improves the reliability of invariance testing.
• The DMI method and accompanying software facilitate more accurate cross-group comparisons.