The heterogeneity in shipper's value of time, results from an SP experiment using mixed logit and latent class

The understanding of the heterogeneity of preferences is taking an increasing role in the analysis of transport behaviour. Such understanding has been developed more extensively regarding passenger transportation than regarding freight transportation. However there are solid reasons, based on empirical evidences or heuristic knowledge, to believe that transportation science should take advantage of the new analytical tools developed in the latest decades to incorporate heterogeneity in the frame work of Random Utility Maximisation. This paper aims at shedding some light on the question of heterogeneity of preferences among shippers. A first section sets up a model to investigate the cause of heterogeneity for a specific transport attribute namely transport time. Then the different methods to estimate heterogeneity are reviewed. In a second section, we implement analytical tool such as Latent Class and Mixed Logit in order to show evidences of heterogeneity. These results are based on preliminary data processing of a survey among shippers in North-East and Central Italy. We also calculate bayesian posterior individual estimates of Parameters for Mixed Logit model and a class membership model for LC analysis