Stated preference survey for estimating passenger transfer penalties: design and application to Madrid

European Transport Research Review

An Open Access Journal

Table 1 Results of the MNL models of utility functions. Pilot survey

Utility function	Attributes		Parameter value	p-value
U (T₀): no-transfer	In-vehicle time	tveh0	−.3573	.0000*
U (T₀): no-transfer	Mode	mode0	−.1547	.6389
U (T₁): one transfer	Constant	constant1	−3.9160	.0053*
	In-vehicle time in first vehicle	tveh0	−.3705	.0001*
	First mode	mode0	.6236	.0961
	Walking time	twalk1	−.3407	.0010*
	Waiting time	twait1	−.3320	.0009*
	Difference in level	stair1	−.4515	.1113
	In-vehicle time in last vehicle	tveh1	−.3448	.0008*
	Last mode	mode1	.3647	.3869
	Real-time information	info	.1691	.6370
	Crowding	crowd	−1.0272	.0038*
U (T₂): two transfers	Constant	constant2	−5.7962	.0375*
	In-vehicle time in first vehicle	tveh0	−.3255	.0264*
	First mode	mode0	−.0133	.9803
	Walking time in first transfer	twalk1	−.3570	.2880
	Waiting time in first transfer	twait1	−.4379	.0127*
	Difference in level in first transfer	stair1	−1.1385	.0277*
	In-vehicle time in second vehicle	tveh1	−.4372	.0006*
	Second mode	mode1	1.5490	.0059*
	Walking time in second transfer	twalk2	−.9148	.0044*
	Waiting time in second transfer	twait2	−.3712	.0662
	Difference in level in second transfer	stair2	−.7091	.1267
	In-vehicle time in last vehicle	tveh2	−.2790	.0153*
	Last mode	mode2	1.0633	.0605
	Real-time information	info	.3593	.3911
	Crowding	crowd	−2.3642	.0008*