Project Summary
This study analyzed the complex travel behavior of transit users by expanding conventional trip-based approaches by considering full activity-travel tours and patterns as basic units of analysis. A tour was defined as a sequence of trips that begins and ends at home and a pattern was defined as an entire day's sequence of activities and associated travel. We considered basic descriptive analyses to first analyze work tours-the tours that contain at least one work activity-of transit commuters and then used Structural Equation Modeling to identify the factors that determine the work tour choices. Latent Class Analysis (LCA) was then used to describe the pattern behaviors of all transit users. The results obtained using the 2017 National Household Travel Survey dataset suggested that 80 percent of work tours consisted of seven dominant tours and that work tour choice was influenced by a set of socio-demographics, built environment, and activity-travel characteristics. The LCA model suggested that transit users can be divided into five distinct classes, namely regular 9-to-5 commuters, after-work stop commuters, multimodal multiple trip makers, morning non-work travelers, and recurrent transit users, where each class had a representative activity-travel pattern. The results can help transit agencies to identify transit user groups with particular activity patterns and to consider market strategies to address user travel needs and to improve the quality of services provided.
