Hi,
I'm seeking to estimate a simultaneous equation model (perferably with ML/GMM and not 3SLS). But my data exhibits spatial dependencies that have to be controlled for, e.g. by using a spatial error specification. Does anyone know of a package that does this, for either R, Matlab or STATA? I've googled for a while and couldn't find anything, but would rather not code it myself. So if anyone knows of a package that does this (or a trick on how you could use the built-in SEM fuctions of these softwares), I would be very grateful.
Best regards,
Tim
Hi Hans,
thank you for your suggestion. Unfortunately, PROC MIXED isn't really what I'm looking for. For one, I didn't see that it allows for spatial error right out of the box, or did I miss sth. here? Also, I unfortunately don't have any experience nor a license for SAS, but for Matlab and STATA (and since it's free, R). So I really need sth. for either one of these softwares, otherwise I'd rather code it myself than learn yet another scripting language.
So, does anyone have another idea?
Best regards,
Tim
PROC MIXED handles spatial correlations built into the error term. For an example, see Article Beyond Standardized Test Scores: The Impact of a Public Scho...
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Structural Equation Modeling