Bayesian Disaggregate and Aggregate Calibration of Path Logit Choice Models

Table 5

Probabilities for the five representative users (ref. parameters in Tables 1 and 2).

	ID, approach	1, C	2, C	3, B	4, B	5, C	6, C	7, B	8, B
	Data	D	D	D	D	A	A	A	A
	Interviews	124	186	124	186	124	186	124	186
User 103	Time	25.9%	24.5%	25.4%	24.7%	33.9%	33.9%	30.0%	30.1%
	Motorway	53.0%	55.7%	54.2%	55.7%	35.8%	35.8%	44.4%	44.2%
	Cost	21.2%	19.7%	20.4%	19.6%	30.3%	30.3%	25.6%	25.7%
User 102	Time	35.2%	34.0%	35.9%	34.9%	45.0%	45.2%	40.1%	40.4%
	Motorway	50.7%	53.0%	51.6%	52.8%	36.7%	36.7%	43.7%	43.6%
	Cost	14.1%	13.1%	12.5%	12.3%	18.3%	18.1%	16.2%	16.0%
User 59	Time	42.7%	42.7%	43.8%	43.4%	43.9%	44.1%	43.6%	43.8%
	Motorway	45.8%	47.1%	46.3%	47.0%	37.7%	37.7%	41.9%	41.8%
	Cost	11.4%	10.2%	10.0%	9.6%	18.4%	18.2%	14.5%	14.4%
User 88	Time	45.4%	45.0%	46.7%	46.0%	50.7%	50.9%	48.0%	48.3%
	Motorway	47.5%	48.4%	47.9%	48.4%	42.2%	42.2%	44.9%	44.8%
	Cost	7.1%	6.6%	5.4%	5.6%	7.1%	6.9%	7.1%	6.9%
User 87	Time	100.0%	100.0%	99.8%	99.8%	99.9%	99.9%	99.8%	99.8%
	Motorway	—	—	—	—	—	—	—	—
	Cost	0.0%	0.0%	0.2%	0.2%	0.1%	0.1%	0.2%	0.2%

C, Classic; B, Bayesian; D, disaggregate; A, aggregate; the alternative chosen is reported in bold.