modelling stated preferences for discrete choices with case sensitive variables

[Siamak Baradaran]

Hi,

I'm trying to model a discrete choice (4 options available to each respondent in each scenario and each respondent faces 9 different scenarios.
Normally when I model discrete choices with case sensitive variables I use the "asclogit" model which allows two types of independent variables:
* alternative-specific variables: which vary across both cases and alternatives and are specified as characteristics of the choice options, in my case they relate to travel time with different modes of transports, costs, etc, and
* case-specific variables: which only vary across cases, in my case individual respondents and include variables such as gender, age, income, etc.

In case of the stated preference study, my respondents are making 9 different choices so it is like having a panel model. I'm trying to find a general model that can handle both case and alternative specific variables. First I thought that I could do the modelling using "femlogit" module but it seems like the model does not allow for case specific variables as it drops such variables as age, gender, etc which do not vary between the scenarios for each respondents.

is there any model out there than can do the work?

[Phil Bromiley]

You didn't get a quick answer. You'll increase your chances of a useful answer by following the FAQ on asking questions - provide Stata code in code delimiters, readable Stata output, and sample data using dataex.

I don't work in this area so I can't help you substantively.

[Siamak Baradaran]

Hi
I haven' received any responses yet so I try to add some details in hope that my question becomes a bit more clear.

I'm trying to model a choice model (in this case mode choice) from data gathered through a Stated preferences survey. The survey has two main type of questions. ones related to the offered service with each of the mode options in terms of travel time, travel time unreliability and travel cost and socio-demographic variables such as age, gender, income, household composition, vehicle ownership and alike. I can build a multinomial logit model using asclogit module. which allows for both alternative specific options (here in terms of times and costs) and individual specific variables such as socio-demographis information.
In case of the SP material, each individual make 9 choices in 9 different scenarios where each scenarios has 4 options. The only module that I've found so far is femlogit which can handle the mode specific information in terms of costs and time while it cannot handle the demographics which varies across the individuals. I need a tool that can handle both alternvtive specific and demographis,