Understanding Drivers’ Route Choice Under Long-Term Pretrip and Short-Term En-Route Traffic Information Using Generalized Estimating Equations

Abstract

This paper addresses two drivers’ route choice paradigms by modeling the factors that affect drivers’ compliance with a long-term pretrip advised route and modeling drivers’ usage of en-route short-term traffic information. A travel simulator with a real network and real historical congestion levels was used as a data collection tool. A generalized estimating equations (GEEs) model with repeated observations and a binomial probit link function was used to ensure the validity of the statistical analysis. Four different correlation structures were used and compared. The results showed that familiarity with the device that provides the information and severe weather conditions increases the likelihood of complying with the pretrip advised route and following the en-route short-term information. Network familiarity and the number of traffic signals on the pretrip advised route have a negative effect on drivers’ compliance. Providing qualitative information and proximity to the destination increase the usage of en-route traffic information.