Construction of Discrete Choice Models to analyze the evolution of travel behavior over time

Urban mobility systems are undergoing significant transformation due to infrastructural developments, technological innovations, and external disruptions. Understanding how individuals adapt their travel behavior in response to these changes is critical for promoting sustainable transportation. In this context, discrete choice modelling has been widely used to analyze travel behavior and mode choice. However, most existing studies adopt a static perspective, often overlooking the dynamic nature of behavioral change and the role of psychological factors over time. Travel behavior is not only influenced by observable variables such as travel time and cost but also by latent psychological factors, including attitudes, habits, and perceptions. These psychological variables are not fixed; rather, they evolve in response to new experiences, policy interventions, and external shocks. At the same time, past behavior plays a crucial role in shaping current and future decisions, reinforcing habitual patterns or enabling behavioral change. The aim of this dissertation is to investigate the role of psychological variables and habits as past choice in shaping the evolution of travel behavior over time by analyzing a panel dataset collected, using different survey techniques, between 2013 and 2024 in Cagliari, Italy. The study examines several key factors affecting travel behavior, including the introduction of new public transport infrastructure, the shock effects of the COVID-19 pandemic, perceived walking accessibility, and the implementation of a smartphone-based Voluntary Travel Behavior Change Program (VTBC). This dissertation highlights that the introduction of a hard measure, such as the implementation of a new light rail line, plays a significant role in shaping travel behavior, influencing both socio-economic characteristics and psychological attitudes over time. This result underscores the importance of accounting for behavioral persistence and evolving attitudes when evaluating new public transport investments and designing effective sustainable mobility policies. The study further examines the shock effects of the COVID-19 pandemic on travel behavior using panel data collected before, during, and after the pandemic in Cagliari. By applying the Theory of Planned Behavior within a hybrid choice modelling framework, the results indicate an increased intention to use cars, influenced by negative attitudes toward sustainable mobility, perceived behavioral control, and social norms. The research also investigates perceived accessibility as an important determinant of walking behavior. Using longitudinal data collected through the Motiontag application, a passive mobile GPS tracking application, both static and dynamic models are estimated to account for state dependence and serial correlation over time. The results show that perceived walking accessibility and land-use characteristics significantly influence walking decisions, while dynamic models highlight the strong role of habitual behavior in predicting future walking patterns across different temporal levels. Finally, the study evaluates a smartphone-based Voluntary Travel Behavior Change Program aimed at encouraging sustainable mobility through Personalized Travel Plans. The results show that receiving the intervention reduces the probability of car use. In addition to the intervention itself, factors such as car ownership and psychological variables significantly influence both the likelihood of participating in the program and the adoption of more sustainable travel behaviors. Overall, this thesis contributes to the literature by providing a dynamic and behaviorally grounded understanding of travel behavior change, offering important insights for the development of more effective and sustainable urban mobility policies.