Evaluating the reliability of reported distance data in urban travel behaviour analysis

The objective of the present paper is to analyze the accuracy of reported distances in travel behaviour research, and to distil from this analysis some useful recommendations for data collection and handling in activity-based modelling. This issue is important because we know from the literature on distance cognition that conjecture, perception and rounding in distance reporting is a rule, rather than an exception. The outcome is a biased transport modelling result. The paper introduces some theoretical reflections on distance measurements and cognition. Next, using household travel survey data collected in 2000 for the city of Ghent, Belgium, the problem of reported distance reliability is examined. A comparison of travel distances using detour factors revealed that self-reported distances provide reasonable estimates of shortest distance path distances. With respect to the effect outliers and rounding have on the reliability of reported distances it is found that mean detour factors are very much dominated by outliers over short distances, but that rounding has little impact. The accuracy of self-reported distances is also influenced by the socio-demographic profile of the respondents, the characteristics of the trip, and the type of transport mode used.