Public transport route choice modelling: Identification of bias when using smart card data

Abstract

Using Automated Fare Collection (AFC) data for public transport analyses has received much research interest recently, including for estimation of passenger preferences through route choice models. However, an important problem persists since AFC data only includes information about the trip within the public transport system, i.e. stop-to-stop. Not knowing the full trip might lead to estimation bias, especially when estimating route choice models using only the chosen stops. This paper highlights this problem by estimating route choice models based on traditional travel survey data and replicated AFC data. In addition, we propose an improved method in which pseudo origin (destination) points in close vicinity of the actually chosen origin (destination) stops are randomly generated, thus allowing pseudo access and egress to be incorporated. The method notably improves parameter estimates of the route choice model compared to estimation assuming AFC stop-to-stop data. Finally, further improvements to the model are presented.