Accuracy and bias of subjective travel time estimates

Abstract

Travel time is the main factor affecting individuals’ travel-related decisions. Understanding the way people experience and estimate travel time is critical for a better insight regarding travel behavior and consequently for planning transport projects and guiding new policies. Whereas most travel-demand models employ objective time data, the use of subjective time data was proposed to improve model estimation. This study draws on fundamental as well as on recent psychological theories, investigating the discrepancy between subjective and objective travel times. According to one of these theories, the return trip effect, travelers tend to report shorter travel times for return trips compared to outbound trips. In a questionnaire, 174 respondents provided pre-trip time estimates. One group estimated travel times from home to local shopping areas, whereas a second group estimated times of the reverse trips. For comparison, objective times were obtained from Waze, a navigation application providing real-time information. Whereas 48% of time estimates were found accurate, over-estimates were 2.5 times more frequent than under-estimates. A return trip effect was found only for trips to/from poorly familiar shopping areas, highlighting the role of destination familiarity. Interestingly, respondents accurately estimated toll-trips but over-estimated non-toll trips. Presumably, merely thinking about paying the toll led individuals to form expectations of travel time savings in exchange. Linear and non-linear regression models for predicting subjective estimates revealed significant effects for trip frequency, trip direction, destination familiarity, toll-road and gender, amongst other variables. The results offer a fertile basis for incorporating subjective time in demand models.