SCHEDULING BEHAVIOR MODELS OF LONG DISTANCE INTERCITY TRAVEL

摘要

This paper focuses on the scheduling behaviors of long-distance intercity travel. Two factors to consider travelers' perceptions and attitudes on expected schedule delays: subjective probability of schedule delays and safety margin at airport and destination. Mixed logit models are developed to deal with heterogeneity of travelers' attitudes and correlations among schedule alternatives. Main results of this analysis based on SP data can be summarized as follows: (1) subjective probability and safety margin can effectively represent travelers' perceptions on the schedule delays of long-distance travel, (2) mixed logit model is able to alleviate the upward bias of multinomial logit model, and (3) travelers are more sensitive to the safety margin at airport than that at destination.