Understanding the effects of individual attitudes, perceptions, and residential neighborhood types on university commuters’ bicycling decisions

Yujin Park, Gulsah Akar, 2019, in Journal of Transport and Land Use

doi:10.5198/jtlu.2019.1259
Location Ohio
Population Other (specify)
Sample size 1169
Factor analysis type principal components, varimax rotation
Stepwise regression nan
Removal of insignificant variables nan
Reviewed by AR

Abstract

This study investigates the effects of individual perceptions and residential neighborhoods on university commuters’ bicycling decisions using the 2015 Ohio state University Travel Pattern Survey data. We generate eight attitudinal/perceptual components based on the 26 bicycling-related questions that capture detailed perceptions of commuters toward bicycling, neighborhood environments, and residential location choice. We create distinct neighborhood typologies combining land use and socioeconomic characteristics, including population, employment, housing and intersection densities, housing types, median age of housing stock, and median household income. Probit regression models are estimated to assess the effects of sociodemographic, attitudinal/perceptual components and neighborhood types while accounting for the residential self-selection effect. Results show that people residing in different neighborhood types reveal significant attitudinal differences in terms of their conditional willingness to bicycle, and evaluation of bicycle friendliness of neighborhoods and routes. We find that bicyclists are more likely to live in neighborhoods that they perceive as having good-quality for bicycling in terms of access to bicycle facilities and lower traffic levels. Results also show the significant association of neighborhood types with bicycle commuting outcomes. People from medium-density, mixed-use, and suburban single-family neighborhoods are less likely to commute by bicycle as compared to those from highdensity, mixed-use neighborhoods. © 2019 Yujin Park & Gulsah Akar.

Factors

Variable Pattern loading
Bicycle racks are easy to find. (Res preference/perception) 0.9
There are enough parking racks for bicycles. (Convenience) 0.87
Variable Pattern loading
Where I now live is a good neighborhood for bicycling. (Res preference/perception) 0.85
There are off-street bicycle or paved paths in or near my neighborhood that are easy to access. (Res preference/perception) 0.81
There are bike lanes easy to access in my neighborhood. (Res preference/perception) 0.78
There is so much traffic along the street I live on that it would be unpleasant to bicycle in my neighborhood. (Safety) -0.66
Variable Pattern loading
I would not leave my bicycle outside my residence because it might be stolen. (Safety) 0.83
My bicycle might be stolen even if properly secured. (Safety) 0.74
Variable Pattern loading
I would choose to ride a bike if there were more indoor or covered places to store bikes on campus. (Convenience) 0.85
I would choose to ride a bike if there were more bike routes to and from campus. (Safety) 0.8
I would choose to ride a bike if there were options for renting or borrowing bicycles. (Convenience) 0.8
Biking can sometimes be easier for me than driving. (Convenience) 0.65
I try to ride a bike to help improve air quality. (Environmentalism) 0.6
Variable Pattern loading
I can find a place to help repair my bicycle if needed. (Convenience) 0.78
When needed, I can find a convenient place to shower and change clothing after bicycling. (Comfort) 0.78
Variable Pattern loading
Bicyclists do not care about drivers on the road. (Safety) 0.92
Bicyclists do not care about pedestrians on the street . (Safety) 0.91
Variable Pattern loading
Biking reduces environmental impacts. (Environmentalism) 0.89
Biking benefits health and fitness. (Lifestyle) 0.88
Biking gives me the opportunity to save money. (Cost) 0.52
Variable Pattern loading
The roadway conditions on some streets make the route unsafe for bicyclists on my commute route (Safety) 0.84
I would have to take detours from the most direct route in order to use bike paths or bike lanes. (Res preference/perception) 0.83
There are no bicycle lanes or routes near enough for me to ride. (Res preference/perception) 0.72
Variable Pattern loading
Safety in traffic is an important factor. (Safety) 0.91
Safety from crime is an important factor. (Safety) 0.88
Extreme weather conditions are important factors. (Safety) 0.76

Models

Dependent variable Choice of bicycle commute
Model type Binary probit
Sample size 677.0
R2 nan
Adjusted R2
Pseudo R2 (nan) nan
AIC nan
BIC nan
Log-likelihood at zero -196.58
Log-likelihood at constants nan
Log-likelihood at convergence nan
Variable Coefficient p-value
Sociodemographic(Base: Male, Staff) nan nan
Female -0.558 0.0
Faculty 0.648 0.0
Graduate/Professional Student 0.55 0.032
Undergraduate Student 0.555 0.087
Commute Distance(Base: More than 10 miles) nan nan
Less than 1 mile 1.288 0.0
1 to 5 miles 1.357 0.0
5 to 10 miles 0.537 0.016
Neighborhood Types(Base Cluster 1-High-density & Mixed-use) nan nan
Cluster 2: Urban Medium-Density & Mixed-Use -0.56 0.01
Cluster 3: Urban Medium-Density & Mixed-Type Residential -0.243 0.201
Cluster 4: Urban Low-Density & Single Family Residential -0.324 0.187
Cluster 5: Suburban Low-Density & Single Family Residential -0.742 0.0
Constant -1.877 0.0
Dependent variable Choice of bicycle commute
Model type Binary probit
Sample size 677.0
R2 nan
Adjusted R2
Pseudo R2 (nan) nan
AIC nan
BIC nan
Log-likelihood at zero -131.02
Log-likelihood at constants nan
Log-likelihood at convergence nan
Variable Coefficient p-value
Sociodemographic(Base: Male, Staff) nan nan
Female -0.413 0.037
Faculty 0.365 0.226
Graduate/Professional Student 0.617 0.069
Undergraduate Student 0.778 0.091
Commute Distance(Base: More than 10 miles) nan nan
Less than 1 mile 1.36 0.005
1 to 5 miles 1.056 0.003
5 to 10 miles 0.361 0.412
Attitudinal Components about Bicycling nan nan
Conditional Willingness to Use Bicycles 0.812 0.0
Bicycle-Friendliness of Neighborhoods 0.375 0.0
Sensitivity to Safety in Mode Choice -0.303 0.008
Perceived Obstacles to Bicycling on Routes -0.318 0.007
Perceived Additional Benefits of Bicycling 0.803 0.0
Negative Images towards Bicyclists on the Street -0.52 0.0
Availability of Bicycle Racks 0.008 0.936
Concerns about Theft 0.05 0.529
Constant -2.93 0.0
Dependent variable Choice of bicycle commute
Model type Binary probit
Sample size 677.0
R2 nan
Adjusted R2
Pseudo R2 (nan) nan
AIC nan
BIC nan
Log-likelihood at zero -126.42
Log-likelihood at constants nan
Log-likelihood at convergence nan
Variable Coefficient p-value
Sociodemographic(Base: Male, Staff) nan nan
Female -4.64 0.025
Faculty 0.528 0.023
Graduate/Professional Student 0.577 0.091
Undergraduate Student 0.76 0.11
Commute Distance(Base: More than 10 miles) nan nan
Less than 1 mile 1.2 0.015
1 to 5 miles 1.065 0.002
5 to 10 miles 0.407 0.322
Neighborhood Types(Base Cluster 1-High-density & Mixed-use) nan nan
Cluster 2: Urban Medium-Density & Mixed-Use -0.617 0.038
Cluster 3: Urban Medium-Density & Mixed-Type Residential -0.302 0.254
Cluster 4: Urban Low-Density & Single Family Residential -0.536 0.18
Cluster 5: Suburban Low-Density & Single Family Residential -0.876 0.018
Attitudinal Components about Bicycling nan nan
Conditional Willingness to Use Bicycles 0.817 0.0
Bicycle-Friendliness of Neighborhoods 0.367 0.001
Sensitivity to Safety in Mode Choice -0.307 0.003
Perceived Obstacles to Bicycling on Routes -0.322 0.001
Perceived Additional Benefits of Bicycling 0.798 0.0
Negative Images towards Bicyclists on the Street -0.54 0.0
Availability of Bicycle Racks 0.015 0.881
Concerns about Theft 0.004 0.976
Constant -2.602 0.0
Dependent variable Choice of bicycle commute
Model type Binary probit
Sample size 677.0
R2 nan
Adjusted R2
Pseudo R2 (nan) nan
AIC nan
BIC nan
Log-likelihood at zero -117.34
Log-likelihood at constants nan
Log-likelihood at convergence nan
Variable Coefficient p-value
Sociodemographic(Base: Male, Staff) nan nan
Female -0.499 0.023
Faculty 0.424 0.103
Graduate/Professional Student 0.652 0.026
Undergraduate Student 0.732 0.087
Commute Distance(Base: More than 10 miles) nan nan
Less than 1 mile 1.301 0.005
1 to 5 miles 0.881 0.06
5 to 10 miles 0.452 0.39
Neighborhood Types(Base Cluster 1-High-density & Mixed-use) nan nan
Cluster 2: Urban Medium-Density & Mixed-Use -0.56 0.004
Cluster 3: Urban Medium-Density & Mixed-Type Residential -0.214 0.484
Cluster 4: Urban Low-Density & Single Family Residential -0.551 0.075
Cluster 5: Suburban Low-Density & Single Family Residential -0.708 0.013
Residential Self-Selection nan nan
Bicycling conditions were a factor in choosing where I live. 0.35 0.001
Attitudinal Components about Bicycling nan nan
Conditional Willingness to Use Bicycles 0.679 0.0
Bicycle-Friendliness of Neighborhoods 0.223 0.091
Sensitivity to Safety in Mode Choice -0.301 0.009
Perceived Obstacles to Bicycling on Routes -0.254 0.021
Perceived Additional Benefits of Bicycling 0.675 0.0
Negative Images towards Bicyclists on the Street -0.478 0.0
Availability of Bicycle Racks 0.059 0.555
Concerns about Theft -0.033 0.749
Constant -3.379 0.0

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