Skip to main content

Table 4 Mixed logit models of transport mode choice

From: How equitable are the distributions of the physical activity and accessibility benefits of bicycle infrastructure?

  Commuting (1,788 trips) Non-commuting (3,148 trips)
β 95% confidence interval t-statistic β 95% confidence interval t-statistic
Alternative-specific constants
 Bicycle -4.92 -5.62 to -4.21 -13.66 -2.77 -2.99 to -2.55 -24.86
 Public transport -4.70 -5.55 to -3.85 -10.83 -5.59 -6.02 to -5.16 -25.25
 Car -2.79 -3.41 to -2.16 -8.76 -2.61 -2.81 to -2.41 -25.96
Random parameters (specific alternative)a
 Cycleway distance (Bicycle) -1.89 -2.16 to -1.6 -13.15 -2.07 -2.35 to -1.78 -14.24
 Non-cycleway distance (Bicycle) -2.65 -3 to -2.3 -15.05 -3.35 -3.84 to -2.86 -13.37
 CBD (Bicycle) -1.52 -2.06 to -0.98 -5.52 -2.31 -3.18 to -1.44 -5.19
 Time (Walk) -0.41 -0.45 to -0.37 -22.88 -0.37 -0.39 to -0.36 -49.39
 Time (Public transport) -0.44 -0.46 to -0.41 -40.26 -0.32 -0.34 to -0.29 -25.02
 Time (Car) -1.17 -1.23 to -1.10 -36.13 -0.76 -0.79 to -0.73 -47.47
 CBD (Car) -5.66 -6.18 to -5.13 -21.07 -3.71 -4.21 to -3.2 -14.39
Non-random parameters (specific alternative)
 Non-cycleway distance x Low intensity (Bicycle) -1.16 -1.38 to -0.92 -9.68 -0.99 -1.25 to -0.74 -7.76
 Rain > 3 mm (Bicycle)     -0.79 -1.16 to -0.43 -4.26
Model fit statistics
 Log likelihood -2478.7 -1928.8
 Chi-square 2887.5 (p < 0.01) 4878.7 (p < 0.01)
 Degrees of freedom 11 12   
 Pseudo-R2 0.58 0.56   
 Akaike information criterion 2091.9 3881.6   
  1. aAll random parameters have a triangular distribution with spread equal to half the mean