Dear users,
I have a question concerning double sample selection bias.
When estimating a probit model on the question whether an individual has information relevant to the decision to be taken in the 2nd stage, will the following method provide me with unbiased results?
probit information x1 x2 x3
from here i calculate the inverse mills ratio and employ it for the next 2 stages:
heckprob decision2 invmills x1 x2 x3 x4, sel (decision1=invmills x1 x2 x3 x4 x5)
Note that decision2 and decision1 are taken by different agents.
Which other method would be appliable. I am not aware of a triple heckman command.
Thank you!
Jason
*
* 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/
