Citation

Jiang Y, Zhang J. Transp. Res. Rec. 2014; 2413: 74-83.

Copyright

(Copyright © 2014, Transportation Research Board, National Research Council, National Academy of Sciences USA, Publisher SAGE Publishing)

DOI

10.3141/2413-08

PMID

unavailable

Abstract

This study investigated how drivers adapted their behavior to traffic congestion caused by accidents on expressways with the help of dynamic travel information. A large-scale stated preference (SP) survey of 2,500 expressway users in western Japan was conducted in 2012. The SP survey was designed on the basis of information on the trip-making experience and preference for travel information for each respondent captured in a revealed preference survey. The targeted context-dependent real-time travel information included information on accident conditions and the impact of the accident, the amount of time predicted to be needed to clear the congestion presented as a single time point and a time interval, and the alternative travel modes available, for three scenarios based on one of the following decision-making times: before departure, on the way to the expressway, and on the expressway. Analyses based on nested logit models found that the predicted clearance time given as a time interval (rather than a specific time point) influenced adaptation behavior considerably more than did the other information and that the influence became greater as the time of the availability of the information moved from before departure to on the way to the expressway and on the expressway. Other factors common across the three decision-making scenarios were distance to the accident site, information that the accident did not involve fatalities, queue length, information that no traffic regulation was in place, the accuracy of the clearance time, and a trend for a decrease in the queue length. The influences of information on fatalities in the accident, clearance time, trip purpose, and clearance time interval were significantly different across the three decision-making scenarios (i.e., they were context dependent).