Effects of Coverage Area Treatment, Spatial Analysis Unit, and Regression Model on the Results of Station-Level Demand Modeling of Urban Rail Transit

Table 5

Negative binomial regression model outputs.

		Equal division		Naive method		Thiessen polygons
		CBG	CT	CBG	CT	CBG	CT
New York	Adjusted	0.67	0.662	0.652	0.62	0.624	0.640
San Francisco		0.71	0.69	0.713	0.721	0.733	0.723
Chicago		0.441	0.450	0.431	0.462	0.378	0.385
Philadelphia		0.607	0.633	0.594	0.615	0.619	0.628
Boston		0.473	0.438	0.273	0.251	0.428	0.416
New York	MAPE	68.168	69.302	69.302	78.57	80.5	75.684
San Francisco		36.268	36.01	35.252	34.88	34.649	35.743
Chicago		56.809	56.802	59.602	56.513	65.188	63.648
Philadelphia		144.898	126.649	157.06	137.508	137.094	129.996
Boston		156.525	173.617	249.653	270.122	169.63	180.475
New York	AIC	13072.860	13082.100	13082.100	13134.810	13130.290	13111.300
San Francisco		828.829	828.838	827.508	826.25	825.107	826.703
Chicago		4104.980	4101.555	4106.677	4099.243	4119.761	4117.203
Philadelphia		3769.013	3758.913	3774.007	3766.500	3766.702	3761.777
Boston		2390.601	2398.985	2437.039	2440.725	2402.503	2404.642