dynamic : An R Package for Deriving Dynamic Fit Index Cutoffs for Factor Analysis.

Area of Science:

• Psychometrics
• Statistical Modeling
• Quantitative Psychology

Background:

• Confirmatory factor analysis (CFA) model fit is often assessed using indices like SRMR, RMSEA, and CFI.
• Current benchmarks (e.g., SRMR=.08, RMSEA=.06, CFI=.96) by Hu and Bentler are widely used but have limitations.
• Fixed benchmarks are sensitive to model characteristics and may not accurately reflect model fit across different models.

Purpose of the Study:

• To address the limitations of fixed fit index cutoffs in confirmatory factor analysis.
• To provide empirical researchers with an accessible method for calculating dynamic, model-specific fit index cutoffs.
• To improve the accuracy and sensitivity of model fit evaluations in structural equation modeling.

Main Methods:

• Introduction of the R package "dynamic" for automating the computation of dynamic fit index cutoffs.
• The package utilizes custom Monte Carlo simulations tailored to the user's specific model characteristics.
• Requires only a lavaan object as input, simplifying the process for researchers without advanced statistical training.

Main Results:

• The "dynamic" R package successfully automates the generation of customized fit index cutoffs.
• These dynamic cutoffs are designed to be more sensitive to model misfit compared to fixed benchmarks.
• The package democratizes access to advanced psychometric methods for a broader range of empirical researchers.

Conclusions:

• Dynamic fit index cutoffs offer a more accurate and tailored approach to evaluating CFA model fit.
• The "dynamic" R package significantly lowers the technical barrier for implementing these improved evaluation methods.
• Adoption of dynamic cutoffs can lead to more reliable assessments of model adequacy in empirical research.