Accounting for uncertainty in the presentation and valuation of travel time in SP experiments

Abstract

Uncertainty is an important consideration when making decisions which means it is important when valuing travel time reliability in SP experiments. This study makes two contributions. First, it demonstrates how to quantify and communicate the uncertainty of quantities of interest when providing value estimates of travel time and reliability to decision-makers. To this effect Bayesian inference is used to propagate uncertainty of parameter estimates to quantities of interest. We find that in particular the estimate of the value of reliability is uncertain, something which would have been missed, if only point estimates are used in its computation. Second, this study uses a novel approach to represent travel time uncertainty in the SP experiment by using quantile dotplots, which we find respondents to perceive as helpful and understandable.