Research Article

Integrated Model of Joint Residence-Workplace Location Choice and Commute Behavior Using Latent Class and Mixed Logit Methods

Table 4

Estimation results of the LCM.

Latent class1234Variable values

Residence location
 10.0570.4770.0180.000Xicheng
 20.0520.4860.1400.000Dongcheng
 30.0010.0000.0190.131Chongwen
 40.0040.0180.0000.079Xuanwu
 50.4110.0020.0000.000Haidian
 60.3640.0050.8210.001Chaoyang
 70.0510.0040.0020.480Fengtai
 80.0600.0080.0000.309Shijingshan
Workplace
 10.0920.2180.0000.024Xicheng
 20.0920.2250.2050.004Dongcheng
 30.0150.0140.0110.094Chongwen
 40.0160.0370.0000.106Xuanwu
 50.3060.1670.0000.002Haidian
 60.4400.2780.7840.000Chaoyang
 70.0300.0480.0000.469Fengtai
 80.0090.0130.0000.301Shijingshan
Commute distance
 10.2150.0830.3770.332<4
 20.3380.4060.5000.5234–8
 30.1580.2560.0860.1118–12
 40.1380.1250.0230.01212–16
 50.0700.0680.0140.01716–20
 60.0810.0620.0000.005>20
Commute mode
 10.0970.0310.1980.133Walk
 20.3070.2140.4390.524Bicycle
 30.2350.2860.1650.114Bus
 40.0610.0630.0000.000Subway
 50.3000.4060.1990.229Car
Household monthly income
 10.1560.0680.1580.308<2500
 20.2500.1980.2690.3302501–3500
 30.3130.4060.2690.3413501–5500
 40.2810.3280.3050.021>5500

Latent class probability47.55%11.00%21.30%20.15%