I have a model with one exogenous variable (X) which links to multiple mediators (4 variables M1,M2, M3 ans M4) and then both exogenous and multiple mediator variables links to endogenous variable (Y).
When I run Structural model in AMOS, X is found to be insignificant to Y. But due to existence of multiple mediators variables in the model, I used SPSS Process Macro from Hayes to test the indirect effect through all mediators. The results show that direct effect of X on Y is significant. The question here is the contradiction between these two results from two different software on the same path. Can someone assist on the following
is this possible? and if yes what is possible causes?
Thanks
In theory, the use of different statistical tools should not lead to different results. Therefore, I think most likely you are testing different models.
I'm not familiar with the Hayes Macro you are using. However, I would suggest you to see whether the macro are testing the model with all four mediators included or not.
AMOS is one of the statistical package for structural equation modeling meanwhile SPSS macro is not considered for structural equation model technique. The factor score calculated between those approach is not similar. Because the AMOS package manage to handle multiple item simultaneously that is no need to compute mean.
Thanks for the answers
However, I would like to ask more, in such a situation what can I do?
Just curious Dr Herman .. are u having parallel or serial mediators?
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Structural Equation Modeling