A methodology for incorporating revealed preference (rp) and stated preference (sp) data in discrete choice models is presented. Themethodology is applied to intercity travel mode choice analysis. New mode shares for each origin-destination pair resulting from changes in service levels are predicted. The combined estimation techniquewith rp and sp data is developed to promote advantages of the two complementary data sources. The empirical study of intercity travel demand demonstrates the practicality of the methodology by accuratelyreproducing observed aggregate data and by applying a flexible operational prediction method. This paper appears in transportation research record no. 1328, Travel demand forecasting: new methodologies and travel behavior research 1991
Abstract