Exploring attitudinal group differences in preferences for shared e-scooter use and its integration with public transit

With increasing interest in the potential of shared e-scooters to enhance urban travel, this paper examines mode choice behavior related to shared e-scooters and the integration of e-scooter with public transit (i.e., scoot-N-ride). We collected stated choice experiments data from two cities in the US, Washington D.C. and Los Angeles. Attitudinal groups were identified by distinguishing attitudinal factors through exploratory factor analysis (EFA) and applying k-means clustering. Clustering results identify four attitudinal clusters with meaningful heterogeneity: pro-cars & against e-scooters, tech-savvy & pro-cars, e-scooter enthusiasts, and pro-noncar-modes. To investigate sensitivities to e-scooter attributes by attitudinal groups, we applied a multinomial logit model (MNL) with interaction terms, enabling a comparison of the potential for shared e-scooter or scoot-N-ride to serve as substitutes for previously used modes. Using the model results, we computed the value of travel time (VOT) for each group, an indirect measure of people’s sensitivity to travel time. By definition, VOT was derived from the coefficients of travel time and cost to assess sensitivities for shared e-scooter and scoot-N-ride. We also conducted mode-specific scenario analysis based on travel distance using the model. Through the analysis process, we were able to specifically identify groups with favorable, moderate, and unfavorable attitudes toward shared e-scooter and scoot-N-ride. These findings contribute to identifying people’s attitudes toward shared e-scooter or scoot-N-ride as well as understanding their preferences. From a policy perspective, understanding the characteristics of groups with moderate attitudes toward e-scooters and adopting tailored strategies could improve e-scooter operations, making them more favorable toward e-scooter adoption.