Some practical issues of demand forecasting of new urban railroad utilizing RP and SP data

Abstract

This study investigated the applicability of the demand forecasting models estimated from revealed preference and stated preference data, focusing on the practical issues such as the reliability of stated intention, treatment of the inertiaterm and the alternative-specific constants of new services, and survey design of SP experiment A panel data, which was collected in Nagoya city before and after a new subway service was started, was utilized to evaluate the predictive accuracy of the models. The results showed that the predictive accuracy highly depends on the treatment of stated intention and the interpretation of the inertia term, and may produce the serious biases in demand forecasting when discrepancy between stated intention and actual behavior was high. On the other hand, repeatedchoice data between existing alternatives with various hypothetical settings might be worth utilizing since it significantly improve the statistical efficiency of the estimated parameters and predictive accuracy.