Influences of Neighborhood Characteristics and Personal Attitudes on University Commuters’ Public Transit Use

Namgung, Mi and Akar, Gulsah, 2015, in Transportation Research Record: Journal of the Transportation Research Board

Location	Ohio State University
Population	Other (specify)
Sample size	2527
Factor analysis type	principal components, varimax rotation
Stepwise regression	no
Removal of insignificant variables	no
Reviewed by	AR

Abstract

This study examined the links between attitudes, the built environment, and travel behavior on the basis of data from the Ohio State University's 2012 Campus Transportation Survey. The analysis results indicated that attitudes might have explained travel behavior better than the built environment. Survey respondents were asked questions about their attitudes on public transit use, and their answers were grouped into new attitudinal factors by using principal component analysis. Then, new neighborhood categories were created by K-means cluster analysis by means of built-environment and land use variables (population density, employment density, housing density, median age of structures, percentage of single-family housing, and intersection density). As a result of this analysis, discrete neighborhood categories, such as urban high-density and residential neighborhoods, and urban low-density and mixed-use neighborhoods, were created. Then, differences in attitudes toward public transit were analyzed across these new neighborhood categories. Binary logit models were estimated to determine the influence of these neighborhood categories as well as personal attitudes on public transit use after sociodemographic characteristics were controlled for. The results indicated that attitudes were more strongly associated with travel behavior than with neighborhood characteristics. The findings of this study will aid in the formation of a better understanding of public transit use by highlighting the effects of attitudes and neighborhood characteristics in transit use as well as differences in attitudes between neighborhood types.

Factors

Variable	Pattern loading
I read books or do other stuff on the bus (Convenience)	0.693
I enjoy spending time on the bus reading or listening to music (Convenience)	0.685

Variable	Pattern loading
I often make other trips during commute to campus (Convenience)	0.561
I often change daily travel plans (Convenience)	0.486
Saving travel time more important than saving money (Convenience)	0.458
Taking bus lengthens commute (convenience)	0.435

Variable	Pattern loading
Transit service is available where I live (Res preference/perception)	0.472
I know where to access bus schedules and route information (Convenience)	0.634
It is not a hassle to search for transit-related information (Convenience)	0.544

Variable	Pattern loading
I don't think about travel options; I just get in my car and go (Convenience)	0.518
I have no interest in reducing car use (Mode-liking)	0.484
I enjoy driving alone (Mode-liking)	0.432
Regardless of adverse environmental impacts, I prefer driving (Environmentalism)	0.496

Variable	Pattern loading
Driving on congested roadway is very stressful (Comfort)	0.72
Driving in traffic congestion more stressful than waiting for bus or bus transfers (Comfort)	0.63

Variable	Pattern loading
Safety from crime is important in mode choice (Safety)	0.679
Safety in traffic is important in mode choice (Safety)	0.691

Variable	Pattern loading
Friends usually take public transit (Social norms)	0.557
Family usually takes public transit (Social norms)	0.53
Coworkers usually take public transit (Social norms)	0.586

Variable	Pattern loading
If buses come often enough, I would use bus more often (Convenience)	0.512
If buses arrive on time, I would use bes more often (Convenience)	0.579
If buses were not overcrowded, I would use public transit more often (Comfort)	0.506

Models

Dependent variable	determinants of transit use
Model type	binary logit
Sample size	533.0
R²	nan
Adjusted R²
Pseudo R² (nan)	nan
AIC	nan
BIC	nan
Log-likelihood at zero	-349.4324
Log-likelihood at constants	nan
Log-likelihood at convergence	-309.82894

Variable	Coefficient	p-value
Constant	2.847	0.01
Socioeconomic Variables	nan	nan
Status(Undergraduate student is base case)	nan	nan
Faculty	-1.742	0.0
Staff	-2.178	0.0
Graduate student	-0.966	0.0
Female	0.193	0.332
Ethnicity(non_white)	0.398	0.136
Locations(1-5 mi is base case)	nan	nan
Less than 1 mi	-0.871	0.002
6-10 mi	-0.302	0.285
11-15 mi	-0.37	0.254
More than 15 mi	-0.838	0.05

Dependent variable	determinants of transit use
Model type	binary logit
Sample size	533.0
R²	nan
Adjusted R²
Pseudo R² (nan)	nan
AIC	nan
BIC	nan
Log-likelihood at zero	-359.4324
Log-likelihood at constants	nan
Log-likelihood at convergence	-266.1624

Variable	Coefficient	p-value
Constant	1.722	0.129
Socioeconomic Variables	nan	nan
Status(Undergraduate student is base case)	nan	nan
Faculty	-1.822	0.0
Staff	-2.251	0.0
Graduate student	-1.272	0.0
Female	0.346	0.131
Ethnicity(non_white)	0.346	0.258
Locations(1-5 mi is base case)	nan	nan
Less than 1 mi	-0.937	0.003
6-10 mi	0.174	0.589
11-15 mi	0.499	0.187
More than 15 mi	-0.249	0.617
Preference of car use	-0.196	0.031
Willingness to use transit	0.261	0.003
Need for flexibility and sensitivity to time	-0.273	0.002
Transit use around a traveler	0.163	0.049
Ability to rest or read	0.282	0.002
Perceived availability of transit service and familiarity with bus information access	0.133	0.165
Sensitivity to congestion	0.138	0.156
Sensitivity to safety	-0.211	0.032

Dependent variable	determinants of transit use
Model type	binary logit
Sample size	533.0
R²	nan
Adjusted R²
Pseudo R² (nan)	nan
AIC	nan
BIC	nan
Log-likelihood at zero	-359.4324
Log-likelihood at constants	nan
Log-likelihood at convergence	-306.49029

Variable	Coefficient	p-value
Constant	1.188	0.0
Socioeconomic Variables	nan	nan
Status(Undergraduate student is base case)	nan	nan
Faculty	-1.78	0.0
Staff	-2.241	0.0
Graduate student	-1.001	0.0
Female	0.163	0.418
Ethnicity(non_white)	0.447	0.097
Locations(1-5 mi is base case)	nan	nan
Less than 1 mi	-0.625	0.06
6-10 mi	-0.205	0.522
11-15 mi	-0.105	0.787
More than 15 mi	-0.461	0.368
Clusters	nan	nan
Cluster 2. Urban high-density, old is base case	nan	nan
Cluster 1. Urban high-density, mixed use	-0.741	0.023
Cluster 3. Urban medium-density, residential	-0.682	0.027
Cluster 5. Urban low-density, residential	-0.877	0.037
Cluster 6. Urban low-density, mixed use	-1.113	0.01
Cluster 7. Suburban low-density, single family	-0.991	0.171

Dependent variable	determinants of transit use
Model type	binary logit
Sample size	533.0
R²	nan
Adjusted R²
Pseudo R² (nan)	nan
AIC	nan
BIC	nan
Log-likelihood at zero	-359.4324
Log-likelihood at constants	nan
Log-likelihood at convergence	-261.85128

Variable	Coefficient	p-value
Constant	0.851	0.01
Socioeconomic Variables	nan	nan
Status(Undergraduate student is base case)	nan	nan
Faculty	-1.932	0.0
Staff	-2.375	0.0
Graduate student	-1.378	0.0
Female	0.293	0.204
Ethnicity(non_white)	0.371	0.234
Locations(1-5 mi is base case)	nan	nan
Less than 1 mi	-0.632	0.078
6-10 mi	0.227	0.535
11-15 mi	0.683	0.126
More than 15 mi	-0.003	1.0
Preference of car use	-0.204	0.026
Willingness to use transit	0.252	0.004
Need for flexibility and sensitivity to time	-0.252	0.005
Transit use around a traveler	0.172	0.041
Ability to rest or read	0.297	0.001
Perceived availability of transit service and familiarity with bus information access	0.1	0.298
Sensitivity to congestion	0.159	0.107
Sensitivity to safety	-0.194	0.051
Clusters	nan	nan
Cluster 2. Urban high-density, old is base case	nan	nan
Cluster 1. Urban high-density, mixed use	-0.814	0.023
Cluster 3. Urban medium-density, residential	-0.564	0.112
Cluster 5. Urban low-density, residential	-0.731	0.136
Cluster 6. Urban low-density, mixed use	0.946	0.049
Cluster 7. Suburban low-density, single family	-0.98	0.258