Dear all,
I'm looking to carry out a random-effects probit (as in xtprobit), where I want to constrain one coefficient to 1 (e.g. in xtprobit could do that by using "offset"), and instead I would like to estimate the variance of the normal distribution (as in hetprobit -- using offset in xtprobit keeps the variance at 1 as far as I can tell). Is there any chance I can tweak an existing Stata command to do that? Or maybe MLE is my best bet? Any thoughts would be greatly appreciated!
Thanks in advance!
Anja