How to Display R² for SAR, SEM, SARAR, and SDM Models in Stata?

[kyu sho]

Dear Stata Community,

I hope this message finds you well. I am currently conducting research using spatial econometric models, specifically SAR (Spatial Autoregressive Model), SEM (Spatial Error Model), SARAR (Spatial Autoregressive Combined Model), and SDM (Spatial Durbin Model) in Stata. However, I have encountered a challenge: these models do not seem to provide an R² value directly in the output.

As R² is a crucial measure for the goodness-of-fit of a model, I am looking for any methods, commands, or workarounds that would allow me to obtain and display the R² value for these spatial models in Stata.

I have considered calculating a pseudo R² manually, but I am interested in knowing if there is a more straightforward or recommended approach within the Stata environment. Any guidance, advice, or examples from your experience would be immensely helpful.

Thank you very much for your time and assistance. I greatly appreciate any help you can provide.

[Nick Cox]

https://www.stata.com/support/faqs/s...ics/r-squared/ gives one view. search will show you that it was posted in 2003 (although some references have been updated). The fact that it bears the name of someone not a StataCorp employee is what it is.

That said, a great deal can be said about this. area. My impression is that many experienced researchers would be somewhere from sceptical outwards about a researcher putting a great deal of emphasis on any single scalar figure of merit, even for comparing results for different models and the same data, if only because much depends on model parameterization.

https://projecteuclid.org/journals/e...3-EJS2180.full looks like a major paper, but I've not made time to read it.