Hi, I'm using SEM to test mediation. And when I was testing the partial mediation model, the model fit indices show that CMIN=0, df=0, CFI=1, NFI=1, RMSEA=0.563. Will this result be reasonable and accecptable to use?
I would appreciate that if anyone give suggestions kindly. And any references for learning how to use SEM to test mediation, especially how to do model comparisons.
Thanks!
Hi Jun I don't think it's appropriate in your SEM analysis the adjustment indicators seem to be overadjusted, as it should indicate the value of CM and gl, also the RMSA should be max 0.08 while yours indicates that the error of your analysis is very high. To compare model you can perform factorial invariance or multigroup analysis.
Cristian Ramos-Vera Thank you for your reply. Might it be the reason that I did not operate AMOS well? For each varaible, I just listed the observed variable (the value after computing average value of all construct) without draing laten variables. Should I list the unoberved variables with corresponding latent variables?
Thank you in advance.
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Structural Equation Modeling