Dear community,
I am relatively new to STATA and am struggling at identifying the postestimation command to obtain the predicted probabilities of the selection equation.
I have read the documentation and I am still having a hard time to figure out what exactly to do to compare the results of churdle with those of heckman for the selection equation.
Noting that y is the variable with corner solutions, lny = ln(y) and dy=1 when y>0, here is the basic code I use for the heckman command to obtain the predicted probabilities, heckprob, of the selection equation :
I would like to compare the predicted probabilities for the churdle command, churdleprob
Your guidance would be mostly appreciated.
Cheers!
Marc
I am relatively new to STATA and am struggling at identifying the postestimation command to obtain the predicted probabilities of the selection equation.
I have read the documentation and I am still having a hard time to figure out what exactly to do to compare the results of churdle with those of heckman for the selection equation.
Noting that y is the variable with corner solutions, lny = ln(y) and dy=1 when y>0, here is the basic code I use for the heckman command to obtain the predicted probabilities, heckprob, of the selection equation :
Code:
heckman lny $xlist , select(dy = $xlist) predict heckprob , psel
Code:
churdle exponential y $xlist , select($xlist) predict churdleprob /* this is where I am lost */
Cheers!
Marc