Skip to main content
. Author manuscript; available in PMC: 2017 Jun 27.
Published in final edited form as: Sci Total Environ. 2014 May 21;490:37–43. doi: 10.1016/j.scitotenv.2014.04.111

Table 3.

Generalized linear model results testing the association between bike route type and pollution levels with and without adjusting for traffic density, background levels, intersections and vegetation density.

Parameter BC (ng/m3)
NO2 (ppb)
β p-valuea β p-value
Unadjusted

Bike Path (Reference) 0 NA 0 NA
Bike Lane 682.3 <0.0001 4.80 <0.0001
Designated Bike Lane 305.2 0.569 9.46 <0.0001
Intercept 1673 <0.0001 14.74 <0.0001

Traffic, Background Concentration and Intersections

Bike Path (Reference) 0 NA 0 NA
Bike Lane 580.1 0.0019 4.33 <0.0001
Designated Bike Lane 118.8 0.831 8.98 <0.0001
Traffic Density −0.017 0.206 −8.1×10−5 0.153
Background Concentration 0.095 0.700 0.684 <0.0001
At Intersection −62.43 0.670 1.42 0.020
Intercept 1876 <0.0001 4.18 <0.0001

Traffic, Background Concentration, Intersections and Vegetation Density

Bike Path (Reference) 0 NA 0 NA
Bike Lane 521.3 0.020 1.77 0.117
Designated Bike Lane 52.74 0.927 6.33 0.0052
Traffic Density −0.018 0.188 −1.2×10−4 0.040
Background Concentration 0.101 0.685 0.668 <0.0001
At Intersection −76.32 0.610 1.07 0.083
Vegetation Density −55.77 0.638 −1.73 0.0008
Intercept 1946 <0.0001 7.42 0.0001
a

The probability that the estimate equals zero