This forum is for questions related to the use of Apollo. We will answer some general choice modelling questions too, where appropriate, and time permitting. We cannot answer questions about how to estimate choice models with other software packages.
Before asking a question on the forum, users are kindly requested to follow these steps:
Check that the same issue has not already been addressed in the forum - there is a search tool.
Ensure that the correct syntax has been used. For any function, detailed instructions are available directly in Apollo, e.g. by using ?apollo_mnl for apollo_mnl
To update to the latest official version, just enter install.packages("apollo"). To update to a development version, download the appropriate binary file from http://www.ApolloChoiceModelling.com/code.html, and install the package from file
If the above steps do not resolve the issue, then users should follow these steps when posting a question:
provide full details on the issue, including the entire code and output, including any error messages
posts will not immediately appear on the forum, but will be checked by a moderator first. This may take a day or two at busy times. There is no need to submit the post multiple times.
I'm estimating a multinomial logit model that has many observations per individual (4,303 observations, 46 individuals, 20 alternatives). An extremely simple model (ASCs + one alternative-specific parameter) produces very large robust standard errors. What could be the source of this issue?
Some specifics of the model:
Because there are between 350 and 76 observations per individual, I am working in logs.
I expect most of these parameters to be non-significant, but these standard errors are a lot bigger than what I would expect.
Even though I do get convergence in this model, adding one more alternative-specific parameter produces convergence to a "saddle point," which I thought was impossible for an MNL (shouldn't the likelihood function be concave?)
Could you please share the results and a bit more info on the model and data? When you say an "alternative specific parameter", is that a coefficient for a dummy or a continuous variable? Also, the estimation results would be very helpful.