Hello;

I conducted a discrete choice experiment to study transportation choices. I have 4 alternatives (car, public transport, bicycle, walking) with one of them being the status quo of the participant. Each participant faces 10 choice exercises where he is asked to choose one alternative out of the four, taking into consideration his status quo and the levels of the attributes describing the alternatives. I divided the 10 exercises into two steps. In the first step we have 2 out of the 10 exercises where the alternatives are described only using the time and cost attributes. The 8 remaining exercises describe the alternatives with 4 attributes, the cost, the time and two sanitary attributes presenting the sanitary impact of the alternative.

I recently started using discrete choice models to analyse the data, first, using a multinomial logit model and dividing the data into two sub-datasets (step 1 and step 2). However, I would like to be able to analyse the data as a whole with the 10 choice exercises.

So, I was wondering if there are some precautions I should take when analysing this data with different number of attributes describing the alternatives in the different exercises, especially when specifying the model? or, is there a choice model recommended to be used in this kind of situations?

Thank you in advance for you help!