Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | Tricia D <tricia201401@gmail.com> |
To | statalist@hsphsun2.harvard.edu |
Subject | st: Computing inverse mills ratio from multiple year probit models |
Date | Tue, 7 Jan 2014 14:26:07 +0100 |
Hi Statalisters, I am trying to estimate a heckman sample selection model for panel data. In contrast to the approach suggested in a previous statalist posting ( http://www.stata.com/statalist/archive/2005-06/msg00456.html), I would like to estimate separate probit models for each of my 11 years in the first stage and compute the respective inverse mills ratios. This would be in line with the approached described by Semykina & Wooldridge, 2011, "Estimation of dynamic panel data models with sample selection". However, I am unsure how to proceed from here - How can I compute the final inverse mills ratio to include it into the second stage regression (xtpcse or pooled OLS with time and firm dummies)? I have found a paper using this approach but the authors only state that they computed the inverse mills ratio using the predicted values from the probit models in the different time periods and then included the IMR in the second stage regression. Yet, I am unsure how to compute this in Stata. I am looking for the Stata command to follow this approach - can anyone help with this? Thanks a lot, Tricia * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/faqs/resources/statalist-faq/ * http://www.ats.ucla.edu/stat/stata/