I have a trivial question related to the interpretation of the parameter estimates in a choice model.
The part-worths are calculated on reference value which I guess is 0.
If I have results like, assuming 'Taxes' as baseline
Variable1 [Access] -0.65
Variable1 [Health] 0.12
How do interpret the performance of 'Access' and 'Health'? Is 'Access' is less 'worth' than the baseline and the Health is slightly more?
What puzzles me is that in the probability profiler, 'Taxes' increase the most the probably of choice, followed by 'Health' and last 'Access'.
Why, then Health has a positive estimate over Taxes?