I am trying to estimate a Random Coefficients Logit model using the RSGHB R package. Thought, I came across with 2 main issues:

  1. Why the gDist does not identify a non-normal distribution? No matter what vector I set for gDist the model keeps considering normal distribution for the coefficient. Also, suppose we solve that problem, which of the results are the transformed sample-level mean estimates (mean and st.dev)? There is only the C matrix which contains the individual-level estimates.

  2. As regards the convergence, when can we be positive that the it has been achieved? Moreover, I have a fixed coefficient for ASC which seems not to converge, while other random coefficients do. Also, the F matrix in the results show a significant number of 0 values which creates bias in the ASC mean estimate. When I change this coefficient to random it seems to converge and there are no 0 values. What could be the problem here?


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.