Hi, everyone!
I just received the comments of a reviewer who said:
you conducted EFA and based on EFA results, run the SEM modeling. You are supposed to conduct a CFA to confirm the EFA results and finalize the measurement model before proceeding SEM. You can fairly use half of the sample to test EFA and another half to test CFA.
Actually, in my study, i used the EFA to explore the possible dimensions of the high-order constructs, and then build a PLS-SEM with the results of EFA. However, i don´t think I should do also the CFA.
So, how can I answer the reviewer ? and is my method wrong??
Thanks!!!
It is fine that you've conducted EFA for exploring the dimensionality. The next step is to confirm the dimensionality using CFA. For CFA, ideally, you should use covariance-based SEM (CB-SEM), but you used PLS-SEM. After confirming the factors through CFA, you should run structural model for hypotheses testing (I assume in your case). You can use AMOS software package (very user-friendly interface), for example, for CFA and SEM. You should do each analysis on different samples (e.g. splitting the sample, as the reviewer suggested). If I correctly understand, this is what the reviewer asked you to do.
All the best.
As a reviewer, I always ask the authors to run FA. It is very easy to perform and turns much more robust the solution. EFA provides previous information on cross-loading, wrong-loading, and mainly on no-loading indicators. Removing or reallocating ill-defined indicators prior to run SEM ensures discriminant validity to the model. In Smart-PLS, it is possible to extract factors by the link discriminant validity in the tab Quality. Nonetheless, I recommend using SPSS instead of Smart, due to the possibility of applying rotation, fi varimax or oblimin. To the best of my knowledge, until now, I did not get how to rotate FA in Smart-PLS. If someone knows how to do it, I’ll be glad to learn more. Best regards and count on me ever.
I have also come across this situation. In one of my papers, I used EFA and then the structural model. Recently received review comments that I should run CFA. The reviewers recommended that if you have a valid scale to measure then you should run CFA must. EFA can only run under the exploration stage only. As many experts have already answered this question, let me ask a little more about this:
1. When reliability and validity can be checked by Cronbach alpha and expert validity, still is it mandatory to run CFA to check validity and reliability?
2. If exploratory research extracts factors with their items and loadings, why do we further need to run CFA?
Looking forward to hearing from experts.
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Structural Equation Modeling