查看文章

The paper presents an investigation on the mode choice behaviour of post-secondary students commuting to school in the city of Toronto. It uses a large-scale dataset collected through a web-based travel diary survey among all students of four universities (seven campuses) in Toronto. Multinomial logit (MNL), nested logit (NL) and cross-nested logit (CNL) models are used for investigating home to school trips mode choices. In terms of goodness-of-fit, the CNL outperforms the MNL and NL model. Furthermore, the proposed CNL model shows fundamental improvements over the MNL and NL models by capturing non-proportional substitution patterns. Empirical models reveal that the mode choice behaviour of female students who travel to downtown campuses differ significantly from female students who travel to suburban campuses. Female students who travel towards downtown are more transit and active mode …