Hello,
I have made a structural model including both formative and reflexive constructs.
I'd like, as far as my data claim it, compare 3 models within exactly the same constructs, for 3 different populations. I have to do so and not put my IV as a moderator because there are not the same items behind the constructs, even if my LVs are the same.
So, I can't do it on AMOS (reflexive) neither on smartPLS3 , do you know a kind of method (and software) in which I can compare those structural models to put ahead by comparison the effect of my IV (expertise level)? I was thinking about M+ and invariance population model comparison but as I am a very beginner in SEM I ask researchers who are used to SEM :-)
Thanks a lot
You can perform the multi-group SEM analysis to determine the invariance between the groups you propose.
https://www.youtube.com/watch?v=vlz284idNpQ&t=15s
Hi Roxane, it is not clear at all to me what exactly you want to compare. Perhaps, you can draw the relevant path diagrams and provide some background about the latent and observed variables.
Two comments beyond that
a) I you can do it in Mpuls (not "M+" :) you can probably do it in AMOS
b) Switching from a SEM framework to a PLS framework fundamentelly changes the nature of your model and modelled variables.
Best,
Holger
As formative constructs do not assume covariance between indicators, therefore the methods such as PLS are better suited to models with formative factors.You can use AMOS for reflective constructs . If your model is combination of formative and reflective constructs then it is better to use PLS.
* Zainudin Awang (2015,p 49) used AMOS for formative and reflective constructs ( Please find the attached file).
** for comparison as already mentioned in this post, you can apply multi-group SEM analysis.
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Structural Equation Modeling