Cycling environmental perception in Beijing – A study of residents’ attitudes towards future cycling and car purchasing

Zhao, Nielsen, Olafsson, Carstensen, Fertner, 2018, in Transport Policy

doi:10.1016/j.tranpol.2018.02.004
Location Beijing, China
Population General
Sample size 1427
Factor analysis type principal components, varimax rotation
Stepwise regression no
Removal of insignificant variables yes
Reviewed by LCM

Abstract

This study focuses on three groups: cyclists, non-cyclists, and non-car owners and examines the significance of the perceived cycling environment, current travel behavior, urban form and socio-demographic variables for the respondents’ attitudes towards future cycling and car purchasing. The paper uses survey data (N = 1427) collected in eight Beijing neighborhoods. The analysis is carried out by means of principal component analysis and multinomial logistic regression analysis. The respondents were generally more positive towards continuing cycling or cycling more in the future than towards car purchasing. The perceived cycling environment was found to be associated with respondents’ attitude towards future cycling and car purchasing. The higher the level of satisfaction with the clarity of cycling space allocation and the higher the agreement with pro-cycling policies, the higher the probability that the respondents will cycle in the future and the lower probability that they will buy a car. Associations with current travel behavior indicate that long everyday travel distances (e.g. 10 km to work or longer) negatively affects the respondents attitude towards their future cycling, whereas short everyday travel distances up to 2 km are positively linked to future cycling prospects. Non-car owners’ attitude to future car ownership is strongly linked to socio-demographic status - low education and low income level groups seems to be most unlikely to take up driving in the future. To encourage people to cycle more and drive less, policy should direct efforts to promoting the clarity of cycling space on the street and strengthen pro-cycling policies. Attention should also be given to stabilizing the current travel modes of non-car users, including promoting the image of cycling, improving the service of walking, cycling, public transport and generally by introducing attractive alternatives to private car ownership. © 2018 Elsevier Ltd

Factors

Variable Loading
How satisfied are you with the clarity of the guiding signs painted on the paths? (Convenience) 0.878
How satisfied are you with the clarity of guiding signs on the streets? (Convenience) 0.849
Variable Loading
How satisfied are you with the setting of traffic lights on prioritizing the cyclist's dimension? (Convenience) 0.788
How satisfied are you with the lighting of the bike paths? (Safety) 0.713
How satisfied are you with the intersection environment for cycling? (Safety) 0.602
Variable Loading
How satisfied are you with the comfort level of cycling in Beijing? (Safety) 0.783
How satisfied are you with the health effect of cycling in Beijing? (Safety) 0.712
How satisfied are you with the time efficiency of cycling in Beijing? (Convenience) 0.541
How satisfied are you with the perceived safety of the cycling? (Safety) 0.49
Variable Loading
How much do you agree that improving cycling can promote the quality of the urban environment? (Transport policy) 0.719
How much do you agree that politicians should give more attention to increasing cycling? (Transport policy) 0.714
How much do you agree that Beijing should revitalize cycling culture? (Transport policy) 0.574
How much do you agree that cycling is a green energy efficient travel mode? (Environmentalism) 0.557
Variable Loading
How satisfied are you with the density of bicycle paths? (Safety) 0.752
How satisfied are you with the physical separation between car lanes and bike lanes? (Safety) 0.697
How satisfied are you with the width of bicycle paths? (Safety) 0.692
How satisfied are you with the continuity of bicycle paths from one to another? (Safety) 0.689
How satisfied are you with the general physical cycling environment in Beijing? (Safety) 0.673

Models

Dependent variable Likelihood of being likely/unlikely to cycle in the future
Model type multinomial logistic regression
Sample size 749.0
R2 nan
Adjusted R2
Pseudo R2 (Nagelkerke) 0.137
AIC nan
BIC nan
Log-likelihood at zero nan
Log-likelihood at constants nan
Log-likelihood at convergence nan
Likely
Variable Coefficient p-value
Intercept 1.376 nan
Number of public facilities within 300 m radius 0.003 nan
Number of bus stops within 300 m radius -0.076 nan
Current travel distance for main trips > 10 m -0.472 <0.05
Clarity of cycling space allocation 0.599 <0.005
Pro-cycling policy 0.585 <0.005
Unlikely
Variable Coefficient p-value
Intercept -1.466 nan
Number of public facilities within a 300 m radius 0.009 nan
Number of bus stops within a 300 m radius 0.013 nan
Current travel distance for main trips > 10 m 0.065 nan
Clarity of cycling space allocation 0.239 nan
Pro-cycling policy -0.163 nan
Dependent variable Likelihood of being likely/unlikely to cycle in the future
Model type multinomial logistic regression
Sample size 552.0
R2 nan
Adjusted R2
Pseudo R2 (Nagelkerke) 0.235
AIC nan
BIC nan
Log-likelihood at zero nan
Log-likelihood at constants nan
Log-likelihood at convergence nan
Likely
Variable Coefficient p-value
Intercept 1.623 nan
Age (years) -0.011 nan
Hukou status (0,1) -0.632 <0.05
Number of public facilities within 300 m radius -26.0 nan
Number of bus stops within 300 m radius -0.038 nan
Current travel distance for main trips <2 km 0.214 nan
Clarity of cycling space allocation 0.001 nan
Pro-cycling policy 0.331 nan
Unlikely
Variable Coefficient p-value
Intercept -0.744 nan
Age (years) 0.047 nan
Hukou status (0,1) 0.128 nan
Number of public facilities within 300 m radius -0.004 nan
Number of bus stops within 300 m radius -0.217 nan
Current travel distance for main trips <2 km -0.557 nan
Clarity of cycling space allocation -0.527 <0.005
Pro-cycling policy -0.15 nan
Dependent variable Likelihood of being likely/unlikely/certain if able to purchase a car in the future
Model type multinomial logistic regression
Sample size 1030.0
R2 nan
Adjusted R2
Pseudo R2 (Nagelkerke) 0.293
AIC nan
BIC nan
Log-likelihood at zero nan
Log-likelihood at constants nan
Log-likelihood at convergence nan
If I obtained the purchasing right, I would buy a car
Variable Coefficient p-value
Intercept 0.59 nan
Age (years) -0.011 nan
Hukou status (0,1) 0.763 <0.005
Driving license (0,1) 0.564 <0.05
Occupation: self-employed 0.731 <0.05
Education: high school or lower -1.115 <0.005
Low income: <1000 yuan/month -0.869 <0.005
Clarity of cycling space allocation -0.138 nan
Pro-cycling policy 0.251 <0.05
Likely
Variable Coefficient p-value
Intercept 0.617 nan
Age (years) -0.019 nan
Hukou status (0,1) -0.101 nan
Driving license (0,1) 0.352 nan
Occupation: self-employed -0.761 nan
Education: high school or lower -0.254 nan
Low income: <1000 yuan/month -1.182 <0.005
Clarity of cycling space allocation 0.033 nan
Pro-cycling policy 0.354 <0.05
Unlikely
Variable Coefficient p-value
Intercept -1.21 nan
Age (years) 0.026 <0.005
Hukou status (0,1) 0.977 <0.005
Driving license (0,1) -0.495 <0.05
Occupation: self-employed -0.307 nan
Education: high school or lower 0.792 <0.005
Low income: <1000 yuan/month 0.488 nan
Clarity of cycling space allocation 0.221 <0.05
Pro-cycling policy 0.183 nan

The Attitudes and Travel Database is produced with support from the Center for Teaching Old Models New Tricks at Arizona State University, a University Transportation Center sponsored by the US Department of Transportation through Grant No. 69A3551747116.

sha256:a08d9e369743bf7e6d1c40d27347318209b40a7fb1543813fdcf31b898918815