st: predict using heckman

[Alfredo Ariza Alfaro <aarialf@upo.es>]

I am using heckman to estimate a wage equation. I would like to impute  
wages to non participants. In order to have predicted wages for non 
participants I can not use some independent variables like tenure, 
experience,etc. cause they are not available for non participants. So, I 
would like to use the variance of the prediction to improve my 
imputations (using two moments of the distribution instead of only one).

I am not sure of what Stata calculate using "predict stdf". Is this the 
variance of the prediction, taking into account the hazard (mills ratio)?

What is the difference between predict stdf  and stdp in Hecman?


Thank you. Regards.


--
Alfredo Ariza Alfaro
Departamento Econom�a y Empresa
Universidad Pablo de Olavide
Carretera Utrera, km.1 
Edificio 3, planta 4, despacho 2
41013 SEVILLA, Spain
telf. + 34 954 34 9187
Fax.  + 34 954 34 9339
http://webdee.upo.es/aarialf

e-mail:aarialf@upo.es

"I disapprove of what you say, but I will
defend to the death your right to say it." 
(The Friends of Voltaire, 1906)

*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/