Route Choice Behaviour: Stated Choices and Simulated Experiences

Abstract

Surveys with stated choice experiments (SCE) are widely used to examine route choice behaviour in hypothetical choice contexts and to derive values of time and reliability for transport project appraisal purposes. In contrast to revealed choices, stated choices do not let participants experience (the consequences of) any attribute, which is one of the reasons why the external validity of SCE outcomes is often questioned. In this paper, we investigate the impact of simulated experiences on behaviour in a route choice context. We recruited 74 people who completed both a typical SCE and an incentive compatible driving simulator experiment (DSE), where the latter required respondents to experience the travel time of their chosen route and actually pay any toll costs associated with the choice of a tolled road. The choices are analysed via a heteroscedastic latent class model. Compared to the SCE, in the DSE, participants selected the tolled road less often, suggesting that having to pay actual money changes stated preferences. Furthermore, we found large variations in sensitivity to toll cost across participants. On the other hand, we found only minor differences in preferences towards travel time and travel time unreliability between SCE and DSE.