I would like to konw how I define the number of elements a sample must have when using SEM ?
I also woul like to know how I can identify that a variable has the same covariance of another whem using the software AMOS 21 when solving SEM problems.
Hi, it depends on the model you want to test (numbers of free parameters, and so on...).
In general there are some rules of thumb (see http://davidakenny.net/cm/fit.htm).
You can also use monte carlo simulation to determine sample size (see www.statmodel.com/download/FinalSEMsingle.pdf).
Regarding your 2nd question: You have to constrain the parameters to be equal and perform a chi²-difference test. But I don't know how to do this in amos, sorry.
Christoph
Accurate answer Christopher...
The link provided is very very interesting and helpful... Thanks a lot...
This can be done in lavaan (http://lavaan.ugent.be) in R.
You are testing two models, one with the covariances allowed to be freely estimated, another with them constrained to be equal. Then test the difference between the two models using the anova function. (which actually compares the two chi squares.)
I think that all SEM-programs allow linear constraints. So I'm sure that you can do this in AMOS. But if you arn't yet familiar with AMOS, lavaan would be a good choice.
Hi friends,
With this help I have solved my problems. Thank you very much.
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Structural Equation Modeling