Exploring preferences for demand-responsive transport in rural Japan: An integrated probabilistic choice set and latent variable approach

Abstract

Although demand-responsive transport (DRT) has been proposed as an innovative mobility solution to enhance accessibility in rural areas, its adoption and utilization remain low. To facilitate broader acceptance, it is essential to identify the factors influencing individuals’ willingness to use such services. This study examined preferences for DRT based on stated preference data collected from 610 respondents in a rural region of Japan. The analysis showed low willingness to use the service, with 75.4% of respondents indicating that they would not utilize it for any trip purpose, regardless of service attributes. To explicitly account for this non-compensatory behavior, an integrated probabilistic choice set and latent variable (IPCSLV) model was developed. Compared with traditional choice models, the IPCSLV model demonstrated a superior fit and yielded more interpretable results. The findings indicated that older adults aged ≥ 75 years, non-drivers, and individuals who were aware of the nearest bus stop were more likely to consider DRT as a transport alternative. Positive perceptions of DRT were also linked to higher levels of acceptance, with participation in the demonstration trial contributing to the development of these favorable views. Among those who considered DRT, service was most preferred for trips to the railway station, and cost emerged as the most influential service-related factor affecting usage. These insights support the design of tailored services that target the specific user groups identified in the analysis. While individuals with high car dependency were more likely to exclude DRT, offering free trial opportunities may encourage them to consider adopting DRT.