ApolloChoiceModelling forum

Hello,
I recently switched to Apollo (from Stata) and really appreciate it! Thank you for sharing this amazing tool.
I have questions related to the dummy coding of variables, and I hope they are relevant to this post.

I have design like this
3 unlabelled alternatives (1 statu quo with fixed levels and 2 other alternatives)
diagnostic: 3 levels 1, 2, 3 (level 1 for statu quo), in the MNL, level 1 would be the based level
Risque1: 3 levels 1, 2, 3 (level 2 for statu quo), in the MNL, level 3 would be the based level
Risque2: 2 levels 1, 2 (level 1 for statu quo), in the MNL, level 2 would be the based level
Cost: 0, 20, 40, 60, 80, 100 (0 for statu quo)

This is the list of utilities
V[['altA']] = b_diag2 * (diagnosticA==2) + b_diag3 * (diagnosticA==3) +
b_risque1_1 * (risque1A==1) + b_risque1_2 * (risque1A==2) +
b_risque2_1 * (risque2A==1)
b_cost * costA

V[['altB']] = b_diag2 * (diagnosticB==2) + b_diag3 * (diagnosticB==3) +
b_risque1_1 * (risque1B==1) + b_risque1_2 * (risque1B==2) +
b_risque2_1 * (risque2B==1)
b_cost * costB

This is the part that I am struggling with: I am not sure how to specify the variable in the statu quo: If I understand well, the attribute "diagnostic" in the statu quo takes the level 1, which is also the based level, so I can omit it, as well as cost attribut (that takes 0 in statu quo). But for the 2 other attributes: they take fixed levels in the statu quo, which are not the based level.
V[['statuquo']] = asc_ref +
b_risque1_1 * risque1_ref + b_risque1_2 * risque1_ref + #### risque1_ref takes fixed value of 2
b_risque2_1 * risque2_ref ##### risque2_ref takes fixed value of 1

I run the code for a simple MNL model, it works, but i am not sure if the results give me what I want.

Thank you very much for your help.
Best regards

Hello again,
I posted a question (not valided yet) but I think I might have figured it out.
Please tell me if I am wrong:
V[['statuquo']] = asc_ref +
b_risque1_2 + ### b_risque1_2 * (column of 1)
b_risque2_1 ### b_risque2_1 * (column of 1)

Thank you in advance,

Hi

yes, that should work fine

Stephane