Dear Statalist,
I want to perform a alternative-specific multinomial probit model with Heckman selection and hoped cmp would allow me to do this. I now have the following command using the travel4.dta dataset, but I don't think it is correct. Would it be possible to do an asmprobit with Heckman selection? And if so, what would it look like?
Kind regards,
Mathilde
I want to perform a alternative-specific multinomial probit model with Heckman selection and hoped cmp would allow me to do this. I now have the following command using the travel4.dta dataset, but I don't think it is correct. Would it be possible to do an asmprobit with Heckman selection? And if so, what would it look like?
Code:
cmp (air:choice1 = t*1) (train:choice2 = psize t*2) (bus:choice3 = psize t*3) (selectvar = psize hinc), indicators(selectvar*$cmp_mprobit selectvar*$cmp_mprobit selectvar*$cmp_mprobit $cmp_probit) ghkanti ghkdraws(200) ghktype(hammersley) constraints(1/2) nodrop technique(dfp nr) nrtolerance(1e-4)
Kind regards,
Mathilde
