External Validity Test for Discrete Choice Transportation Forecasting Models Based on the Stated Choice Approach

摘要: A Constrained Conditional Logit Model (CCLM) based on the hypothesis that existing resource constraints have significant effect an individual's mode switching is developed. Theoretical and empirical studies reported here supported validity of this hypothesis. Data were collected traveler demographics commuter modal choice in response to parking charges at two major urban New Jersey corporate sites. Several developmental contributions advancements are reported, including: (1) development CCLM which explicitly includes constraint into decision making process then a probability function, therefore providing tool for evaluating various constraints; (2) advance model estimation, not only improves efficiency estimation but also provides study transferability models; (3) cross-sectional design makes external test available situation where researchers no opportunity manipulate both control conditions actual experiment. This shows can successfully forecast split change associated with charge imposition single occupancy vehicle public transit modes.