unequal attributes in unlabeled stated choice
Posted: 19 Jan 2022, 04:37
Hello everyone,
I have data from two similar unlabeled experiments, both with two alternatives (A and B).
Experiment 1) three attributes: example(travel time, travel cost, parking cost)
Experiment 2) two attributes: example(travel time, travel cost)
Second experiment have around 20 % of respondents and first has 80 % of respondents.
The question is, is it okay to use this data together in a single model (with same utility specifications) by making parking cost
equal to zero for 20% of the respondents (for both alternatives A and B) or should i use specifications for scaled differences like
with joint RP and SP method, or what is the best way to model for such data.
I have data from two similar unlabeled experiments, both with two alternatives (A and B).
Experiment 1) three attributes: example(travel time, travel cost, parking cost)
Experiment 2) two attributes: example(travel time, travel cost)
Second experiment have around 20 % of respondents and first has 80 % of respondents.
The question is, is it okay to use this data together in a single model (with same utility specifications) by making parking cost
equal to zero for 20% of the respondents (for both alternatives A and B) or should i use specifications for scaled differences like
with joint RP and SP method, or what is the best way to model for such data.