Yes, in JASP , you can

These indices can provide more nuanced insights compared to relying solely on traditional methods like eigenvalues greater than 1 (Kaiser criterion) or visual inspection of the scree plot.

You did not mention sample size.

Common Fit Indices in JASP for EFA

Root Mean Square Error of Approximation (RMSEA):

Lower values (e.g., below 0.06) indicate a better fit.

Comparative Fit Index (CFI):

Compares the fit of the specified model to a baseline (null) model.

Values close to 1 (e.g., above 0.95) suggest good fit.

Tucker-Lewis Index (TLI):

Adjusts the model fit for complexity, similar to CFI.

Values closer to 1 indicate better fit.

Chi-Square Test of Model Fit:

Tests the discrepancy between the observed and reproduced correlation matrices.

A non-significant result suggests a good fit, although it is sensitive to sample size.

Bayesian Information Criterion (BIC):

A criterion for model comparison, penalizing model complexity.

Lower BIC values indicate a better balance of fit and parsimony.

Parallel Analysis:

A method to compare eigenvalues of the observed data with those from randomly generated datasets.

Retain factors with observed eigenvalues greater than those from random data.

Following are Steps in JASP:-

Perform EFA using Principal Axis Factoring in the Factor Analysis module.

Inspect eigenvalues and scree plot for initial guidance.

Look at the additional fit indices provided in the output:

RMSEA, CFI, TLI, and others to evaluate the fit of models with different numbers of factors.

Conduct Parallel Analysis to validate the number of factors to retain.

Use these indices collectively to make an informed decision about the number of factors.

By combining these fit indices with theoretical considerations and the interpretability of the factors, you can achieve a more robust determination of the optimal number of factors to retain.

Hello Riya,

if there is a clear theoretical rationale, why do you use EFA at all? I would suggest testing the model with a CFA. And be prepared that the model will drastically fail. But within failure, there is a chance for you to re-build the scale, but this will probably be limited to 2-3 crystal clear indicators of each of the factors (which is absolutely sufficient).

Here are a few papers that may guide you (focusing on published papers, that tested critically measurement models) and failed:

Hayduk, L. A. (1985). Personal space: The conceptual and measurement implications of structural equation models. Canadian Journal of Behavioural Science, 17(2), 140-149.

Hayduk, L. A., & Beckie, T. M. (1997). Measuring quality of life. Social Indicators, 42, 21-39. https://doi.org/doi.org/10.1023/A:1006881931793

Hayduk, L. A., Pazderka-Robinson, H., Cummings, G. G., Levers, M.-J. D., & Beres, M. A. (2005). Structural equation model testing and the quality of natural killer cell activity measurements. Medical Research Methodology, 5(1), 1-9.

Kline, R, B. (2024). How to evaluate local fit (residuals) in large structural equation models. International Journal of Psychology, 59(6), 1293–1306. https://doi.org/10.1002/ijop.13252

Rosman, T., Kerwer, M., Steinmetz, H., Chasiotis, A., Wedderhoff, O., Betsch, C., & Bosnjak, M. (2021). Will COVID‐19‐related economic worries superimpose health worries, reducing nonpharmaceutical intervention acceptance in Germany? A prospective pre‐registered study. International Journal of Psychology. https://doi.org/doi.org/10.1002/ijop.12753

Hope that helps,

Holger