Title

A JOINT MODE-TRANSIT SERVICE CHOICE MODEL INCORPORATING THE EFFECT OF REGIONAL TRANSPORT SERVICE TIMETABLES

Document Type

Journal Article

Publication Date

2003

Subject Area

ridership - mode choice, mode - bus, mode - rail, mode - mass transit

Keywords

Utility theory, Travel behavior, Transit, Timetables, Schedules, Railroad travel, Public transit, Mode choice, Modal choice, Mass transit, Local transit, Intercity travel, Goodness of fit, Correlation analysis, Correlation (Mathematics), Choice of transportation, Choice models, Bus usage, Bus travel, Automobile use, Automobile usage, Automobile travel, Arrivals and departures

Abstract

This paper proposes a joint mode-transit service choice model that recognizes that the different services offered by transit systems in urban and inter-urban areas lead to different assumptions about mode choice behavior. Most inter-urban mode choice models simulate the choice between private and transit modes based on service frequency. This model assumes that a user is likely to choose a transit mode if a service trip sufficiently compatible with the desired arrival time is available rather than because of its daily frequency. In the proposed model, the elementary alternatives are specific service trips of the available transit modes (bus and train) together with private automobile. Differences between the desired arrival time of the traveler and the arrival time of the specific service trip (early/late arrival penalty) are included rather than the frequency of the relative service/line. Different utility function specifications were tested with both nested and cross-nested correlation structures. Results showed the importance of capturing the complex correlations existing among the alternatives of these choice contexts. A generally high significance of early/late arrival penalty parameters was observed, indicating the importance of such factors in inter-urban mode choice modeling. A comparison of the proposed model with a classic mode choice model showed a significant increase in the goodness of fit of the proposed model despite the large number of alternatives presented.

Comments

Transportation Research Part B Home Page: http://www.sciencedirect.com/science/journal/01912615