Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | Zhi Su <su.zh@husky.neu.edu> |
To | statalist <statalist@hsphsun2.harvard.edu> |
Subject | st: multivariate probit models with unobserved selection |
Date | Sun, 15 May 2011 22:55:30 -0400 |
I have four latent variables y*_1=x_1*b_1+e_1, y_1=1 if y*_1>0 y*_2=x_2*b_2+e_2, y_2=1 if y*_2>0 y*_3=x_3*b_3+e_3, y_3=1 if y*_3>0 y*_4=x_4*b_4+e_4, y_4=1 if y*_4>0 where e_1, e_2, e_3, e_4 are error terms distributed as multivariate normal with variance-covariance matrix V. If a generalized term e is correlated with y*_1, y*_2, y*_3, and y*_4, how to estimate this generalized term? Or a term similar to the inverse Mills ratio in the first stage equation of Heckman 2-stage selection model. -- Zhi Su 348 Holmes Hall Northeastern University 360 Huntington Avenue Boston, MA 02115 Office:1-617-373-2316 email:su.zh@husky.neu.edu * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/