Vuong Test

[fernanda mazzotta]

Dear All,
I am interested in how to perform Vuong's test to choose the best model, between two simultaneously system of three equation estimated with cmp: [1 system]: cragg, tobit and ols, [2] tobit, tobit and ols. Somebody could help me? In a previous message I read of a code for comparing probit and logit, maybe having it can help me. Thank you very much in advance

[Marcos Almeida]

Hello Fernanda,


As far as I can envisage up to this point, the Vuong statistic is applied in count models (negative binomial and Poisson regression) so as to test whether a zero-inflated model is "favoured" if compared to a standard model.

Best,

Marcos

[fernanda mazzotta]

Dear Marcos, thank you for your reply,
I read about using the vuong test also for compare nested and non nested model
http://www.stata.com/statalist/archi.../msg01261.html
and also in a previous chat http://www.statalist.org/forums/foru...nnested-models
the prof Wooldridge that he has "Stata code comparing probit and logit, but it's easy to apply to any case where you can compute the log likelihoods for each observation. It is not enough to compute the total log likelihood"
best
f

[sabrina mines]

Dear Fernanda,
Have you solve the problem ? I am in the same situation and I m looking to the procedure to use Vuong test to compare two part model and Heckman model?
Best