SummaryThis study develops a random effect panel data logit model that identifies the factors that influence expressway users' decision behavior under a new arterial road variable message sign information service. The new information service provides travel time of both an expressway route and an alternate arterial road route. It is based on the data collected from a stated preference survey of Shanghai drivers. Correlations within repeated choices by the same respondent were addressed. The results show that the random effect model performs well in addressing repeated observations and evidences the existence of common unobserved random factors affecting route choice behaviors of the same respondent. It is shown that drivers' decisions can be significantly influenced by the new information service. Driving experience, travel time saving, and occurrence of expressway accidents serve as positive factors whereas number of traffic lights on the arterial road serves as a negative factor in choosing the arterial road. Private car drivers, employer‐provided car drivers, and taxi drivers value number of traffic lights in a different way. Female drivers are more sensitive to expressway delays. Drivers with rich driving experience and female drivers are more sensitive to number of traffic lights. Copyright © 2014 John Wiley & Sons, Ltd.
Read full abstract