Hello everyone,
I am trying to run a discrete choice analysis of cross-sectional data (long format). It is related to entrepreneurs choosing to incorporate their startups in a particular country. There are 13 different options that entrepreneurs have.

My base model works fine. However, once I try to add some of my control variables, I start getting the following message:

cmclogit choice Innovation Risk Fraud Information Consumer Rule VCavailability
note: 36 cases (504 obs) dropped due to no positive outcome per case
note: variable Innovation has 57 cases that are not alternative-specific; there is no within-case variability
note: variable Risk has 7 cases that are not alternative-specific; there is no within-case variability
note: variable Fraud has 57 cases that are not alternative-specific; there is no within-case variability
note: variable Information has 7 cases that are not alternative-specific; there is no within-case variability
note: variable Consumer has 57 cases that are not alternative-specific; there is no within-case variability
note: variable Rule has 57 cases that are not alternative-specific; there is no within-case variability

Iteration 0: log likelihood = -3969.4013 (not concave)
....
Iteration 61: log likelihood = -3957.7521 (not concave)
dimension of beta incorrect
r(503);

I wonder what I am supposed to correct this mistake.

Thank you!

Denis Iurchenko
Cal Poly Pomona