Doctor of Philosophy (PhD)
Geography and Anthropology
Background: Surveys across the U.S. reveal that commuters driving personal vehicles spend a significant amount of time in traffic, while public transit, as an efficient commuting mode, has been largely underutilized.
Purpose: What causes a low public transit ridership? How could public transit ridership be explained by demographic, socio-economic and spatial characteristics of neighborhood? This study answers these questions by deciphering the relationships between public transit ridership and various factors in a medium-size city in southern U.S. – Baton Rouge, Louisiana.
Methods: Non-spatial and spatial data in a larger areal unit (e.g., block group) are used to infer demographic, socio-economic and spatial variables in a smaller areal unit (e.g., census block) to gain a sharper spatial resolution in the analysis of public transit ridership in geographic information systems (GIS). First, the ecological inference method is used to disaggregate demographic and socio-economic data from the block group level to the census block level. Secondly, Monte Carlo simulation and transit schedule data are used to improve the estimation of travel time by private vehicle and public transit, respectively, based on which commuting time ratio of these two is calibrated at the census block level. Regression analyses including ordinary least square (OLS) regression, geographically-weighted regression (GWR) and semi-parametric GWR (SGWR) are used to explain the variability of public transit ridership using demographic, socio-economic, and spatial variables at the census block level.
Results: A stepwise regression process selects six variables from 25 original variables representing different aspects of demographic, socio-economic, and spatial characteristics at the census block level. The final model includes both global and localized effects on public transit ridership. Recent immigrants and carless population are positively related to public transit ridership. White population concentration is negatively related to public transit ridership. These relationships are found to be consistent across the study area. The relationships between public transit ridership and income, commuting time ratio, and accessibility to employment via public transit vary across the study area, and some of these variables even show opposite effects in specific pockets in contrast to their area-wide average effects.
Kuai, Xuan, "Deciphering Public Transit Ridership in Baton Rouge: Spatial Disaggregation Approaches" (2019). LSU Doctoral Dissertations. 5030.