A Geographically Weighted Regression Approach to Modeling the Determinants of On-Demand Ride Services for Elderly and Disabled
Muhammad Arif Khan, Amir Shahmoradi, Roya Etminani-Ghasrodashti, Sharareh Kermanshachi, Jay Michael Rosenberger
Abstract
The number of disabled and senior citizens in the United States has been rising in the recent past. Accordingly, many cities provide on-demand door-to-door paratransit ride services to improve the mobility of this population. Few studies have investigated the ridership trends in small cities where on-demand ride services are the only available transportation option. This study explores the determinants of ridership for the paratransit service in the City of Arlington, TX, using Handitran ridership data. We find that the shares of the male population, senior citizens, vehicle ownership, and average household size are significant predictors of the ridership at the geographical block group level. The aforementioned predictors’ predictive power varies significantly in different city blocks, indicating diverse needs for different areas. Our findings also provide new insights into the implementation of a wheelchair-accessible autonomous vehicle fleet designed to expand accessibility to people with limited mobility.