I am somewhat lost regarding the estimation of a constrained clogit
model in Stata 10. The specification is very simple like, for instance:
clogit dep indep_a indep_b, group(id)
So far not problem at all, but I would like to impose two constraints:
1. The coefficients on indep_a and indep_b should add up to unity. (So,
the model I am really interested in would be something like pr(choice_i
= j) = beta1*ut1 + (1-beta1)*ut2.)
2. The coefficnent(s) should also fall into the inverval from 0 to 1.
Of course, I could implement constraint 1 by applying a simple linear
constraint, but given the data at hand, my results are often out of the
bounds defined in constraint 2.
So, how can I implements constraints 1 and 2 at the same time? Of
course, some kind of a logistic constraint would be a good idea, but how
do I do that in Stata. Any ideas?
