Hello,
I have a question about the calculation of mediator analysis. As we all know, there are different ways to examine mediator variable. One of the methods is using structural equation modelling, as I know. But we get a criticism about this method. Is not it possible to examine mediator effect by structural equation modelling?
Thank you in advance.
Meryem
Hi, Meryem
Could you be a bit more specific about the criticism you noted that relates to SEM? Very often the problem we face when running mediation analysis boils down to an inadequate data (e.g. coming from a cross-sectional study). The SEM approach in itself seems to be quite reliable in my view.
Best,
Lukasz
Dear Lukasz,
The comment of the reviewer is below:
The main shortcoming of this study is that the title of the study is about the mediating role of school culture, while nothing about testing mediation has been reported here. it seems authors had no idea of how to test mediation. To test mediation, some specific statistical procedures are used such as bootstrapping or Sobel test. Therefore, all these mediations should be tested and reported with these specific statistical procedures and one cannot claim mediation without running these analyses.
Is not it possible to test mediation with SEM?
Meryem
Find below my paper for reviewer comments:
Article Making Sense of Mediating Analysis: A Marketing Perspective
Meryem Berrin Bulut are you using PLS SEM or CB SEM? In PLS SEM mediation is done by bootstrapping
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Structural Equation Modeling