Dear all,
I want to estimate an ordered probit model conditional on a selection equation.
I have two problems.
My first method has been to try the ?ssm? command. But even after 36 hours
there I have received no results. Is this reasonable given that I have 18000
individuals, 10 variables in the first equation and 8 variables in the second??
Second, I want to change strategy to use an ordered probit with a Mills ratio,
included from a bivariate probit for the selection equation. I did read the
reply by Maarten Buis on the 5th of October "Inverse mills ratio and ordered
logit " and hence prefer to use this method. But what does this procedure imply
for the error terms and can it create a bias? Does the error terms have to be
corrected and, if so, how?
I am very grateful for any help,
Jenny
*
* 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/
