R-Squared in Tobit Panel

[Mona Wenzel]

Hello everybody,

I am performing a tobit regression on panel data and intend to examine the explanatory power of the model. However, xttobit does not yield an R-squared or an Pseudo R-Squared.
The installation of packages like fitstat and spost 13 neither worked out. After installing them and do fitstat, I get the error message that fitstat does not work with the last model estimated.

Please, can anybody tell me how to obtain the R-Squared value from a tobit panel regression?

Thanks a lot in advance.

Mona

[Marcos Almeida]

Hello, Mona,

Maybe that's not a good idea.

Please read this:http://www.stata.com/support/faqs/statistics/pseudo-r2/

And this: http://www.ats.ucla.edu/stat/stata/o...tata_Tobit.htm

Hopefully that helps.

Best,

Marcos

[Mona Wenzel]

Thank you Marcos, and sorry for the duplicates. Stata issued "unexpected token e" as error when posting, so I assumed that it did not work.

Would you have any other suggestions how I can compare the explanatory power of two tobit panel models? Stating that one model fits better than the other just because of higher coefficients neither appears to me as a sound approach.

Thank you again.

Mona

[Marcos Almeida]

Hello Mona,

I don

[Marcos Almeida]

Hello Mona,

I don't have much experience with xttobit, but the same principles prevail when it comes to modeling: check the CI, predictions, margins and marginsplot. Also, - estat ic -.

Here you have the main postestimation commands for xttobit:http://www.stata.com/manuals13/xtxtt...estimation.pdf

Best,

Marcos