Using appropriate prior information to eliminate choice sets with a dominant alternative from D-efficient designs

Discrete choice experiments are frequently applied in transportation research to explore the value of travel time savings. Most profile attributes in these studies, such as the travel time or the travel cost of a mode or road choice, have a clear rank order in their attribute levels. Therefore one option in a choice set can dominate the other alternatives in the set. This research shows, for a specific setup in the transportation field, that a proper choice of prior information, adequately incorporating the dominance of preferred attribute levels, eliminates choice sets with a dominant alternative from Bayesian D-efficient designs.