In the statistical analysis of the structural equation model (SEM), if we have an independent variable and a dependent variable and two variables as mediators, if we want to calculate some demographic characteristics such as gender and employment status and marriage as a moderator, Amos and, Lisrel software are better or R software?
These programs should all provide identical estimates (within rounding errors) for SEM provided the same estimation method (e.g. Maximum likelihood) and convergence criteria are used.
I should add that the advantage of R (lavaan package) over the other programs is that it is free software.
@Ahadi AMOS won't allow to add moderator when running the SEM model with latent variables. Lisrel and Lavaan in R can handle any kind of model. I totally agree with @Christian that Lavaan has the advantage of being license free but both of these softwares are command based. If you are looking something that can handle your model and is a GUI based software, then you can turn to Smart-PLS. It has a trial version of one month with full features.
Thanks a lot for the complete and useful answers of both dignitaries(@ Dr. Imran Anwar and Dr.Christian Geiser).
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Structural Equation Modeling