Through preliminary studies, I have selected seven attributes for SP modelling. Since I have to make separate models for A.C and Non-A.C buses, I have considered 'type of bus' as one attribute (levels are A.C and Non-A.C). However, while incorporating the attribute ‘fare’ I need to give due consideration to the present fare of A.C (ie. Rs 3.0/km)) and Non-A.C buses (Rs 1.0/km). Therefore, levels of fare considered for the experiment are 1.0, 1.5, 2.0, 2.5, 3.0, and 4.0 (Rs/km.
If I do the D-optimal design, I may get an A.C bus type with a fare less than present fare of A.C bus (say Rs 1.5/km) which is not practical. How can I tackle this issue?
Yes, as Kun Gao suggested, you can consider different set of fare levels and constrained them with bus type. Another issue, as I understand, you are considering an unlabelled experiment and AC and Non-AC bus as attribute... These two bus services are distinctly different from each other, therefore, I would suggest you to go for labelled experiment.
Thanks Kun Gao and Saurabh for your valuable suggestions
- Badges
- Science topic
- Similar topics
- Engineering
- Civil Engineering
- Transportation Engineering