A model has been developed to allocate peak-period and off-peak trips across the Hudson River in New York by car, bus and train. Determinants of route choice were time, cost and convenience, all based on a rating system. Multiple regression analysis on the origin- destination analysis based on comparative usage of the facilities was applied to determine the variations. Results show the at most importance of time for the car driver, suggesting the general validity of the "all-or-nothing" minimum time approach. For the trail user the time factor is not so important, but here is the number chances an important determinant. Results of the model for the base year 1964 gave good agreement with the actual travel volume, although some fine-tuning was required for the transport by auto. Research is continuing with new data.
Abstract