I haven't found an answer to this online: Is there any package / easy to implement way to estimate a probit with heckman correction in R? The corresponding function in STATA would be heckprob

  • $\begingroup$ you may wanna take a look at the sample selection package. See here: cran.r-project.org/web/packages/sampleSelection/vignettes/… and here: cran.r-project.org/web/packages/sampleSelection/… $\endgroup$
    – Sascha
    Mar 13, 2018 at 16:15
  • $\begingroup$ Hi, thanks for the comment :-) I did look at the sample selection package, but still cant see an application for heckman probit :-( $\endgroup$
    – user469216
    Mar 13, 2018 at 16:28
  • $\begingroup$ I see. Here's a practical guide and implementation using the sample selection package: rpubs.com/wsundstrom/t_selection $\endgroup$
    – Sascha
    Mar 13, 2018 at 16:39
  • $\begingroup$ thanks for the link :-) still, the regression of interest is OLS (the 2nd step), not probit :(. $\endgroup$
    – user469216
    Mar 13, 2018 at 16:48
  • $\begingroup$ You need to post a minimal complete verifiable example (MCVE). I suggest linking to a relevant, easily-obtainable data set (would the Smoke data in sampleSelection do?), and reporting the Stata output in your question. Then someone can help you get the same output in R. $\endgroup$ Sep 18, 2018 at 17:02


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.