Transit services and user satisfaction: Application of latent class cluster analysis

Past studies have focused mainly focused on studying different aspects of traditional flexible and fixed route transit services, but little attention has been paid towards ridesharing services available to university community. To bridge this gap, this study is aimed at classifying university community based on their satisfaction levels towards demand responsive transit services available to them using a Latent Class Cluster Analysis (LCCA) approach. We employ LCCA models to find out the clusters of users based on their perceptions towards several service performance attributes of three ridesharing services that serve in the University of Texas at Arlington community. Results show that younger, women and low-income populations are more likely to be satisfied as compared to older, men and high-income populations. We also find that white and domestic students are more likely to be satisfied than Asian and international students. Respondents from households without a vehicle were also more likely to be satisfied than users with more than one vehicle in the household. Findings from this study could be used to understand how different groups of users perceive the service performances of these ridesharing services and this could help transportation planners and services providers to improve the efficiency of their services.