How to get Pseudo R2 for Heckman two stage model?

[Yi Liu]

Stata does not provide Pseudo R² when I run hackman two stage model, could anyone help me how to generate r square or pseudo r square? the current code I have:

/*two steps*/
heckman donat_amount age_diff education_diff , select (donate=age_diff education_diff age_diff education_diff ) twostep
/* default prediction linear predictor*/
margins,dydx(*)
/*marginal effect for probit model can use in Heckman -heckman postestimation*/
margins,dydx(*) predict (psel)

if I want to get marginal effect for both of the stage, am I right for the code above? Also, if i cannot use svy, how can i apply weights on Heckman model?

Thank you!

[Yi Liu]

Anyone knows ? i look forward to some answers. thanks ~

[Carlo Lazzaro]

Yi:
your query was left unreplied mainly because it came with no data excerpt/example, so that interested listers can try to replicate what you're after.
That said, I try to reply to some of the points you raised:
- -heckman- gives back squared chi, not (pseudo)R;
- from -help heckman- it seems that you can add weights.

[Yi Liu]

Yi:
your query was left unreplied mainly because it came with no data excerpt/example, so that interested listers can try to replicate what you're after.
That said, I try to reply to some of the points you raised:
- -heckman- gives back squared chi, not (pseudo)R;
- from -help heckman- it seems that you can add weights.

Thank you for your reply. This is my first posting.
Do you know how to add weights using heckman? Do you know the code for using weights? thank you!

[David Radwin]

The pseudo R-squared question is a FAQ. See these links:
https://www.stata.com/statalist/arch.../msg01186.html
https://www.stata.com/statalist/arch.../msg01187.html

As for weights,

help heckman