Abstract

Recent research has indicated convincing evidence of a link between characteristics of the built environment and travel behavior. However, few land use—travel behavior studies include cognitive factors (such as attitudes, perceptions, and environmental norms) that have been found to affect travel mode choice in the social psychology literature. This dissertation develops and empirically tests a theoretical framework called the Perception-Intention-Adaptation (PIA) model that brings land use and attitude-behavior theory together in order to address gaps in the travel behavior literature. Following a detailed description of the PIA model, the dissertation is comprised of three empirical essays. The analyses in these essays are based on cross-sectional and panel data collected during the Expo Line Study, the first experimental-control, before-and-after evaluation of a rail transit investment in California. The first essay evaluates the predictive power of the core socio-psychological constructs of the PIA (attitudes, norms, and control beliefs) in combination with a comprehensive set of built environment and socio-economic measures. Regression models of transit use are used to analyze cross-sectional data obtained before the opening of the Exposition light rail line in Los Angeles. The analysis indicates that two PIA constructs, attitudes toward public transportation and concerns about personal safety, significantly improve the model fit and were robust predictors of transit use, independent of built environment factors. The second essay uses panel data collected before and after the opening of the Exposition light rail line to examine changes in travel behavior. A quasi-experimental approach with experimental (within ½ mile of an Expo station) and control (beyond ½ mile) households is used to evaluate the travel effects of the opening of the Expo line at the household level. The results show a statistically significant reduction in vehicle miles traveled (VMT) in the experimental group, though overall transit ridership and travel-related physical activity did not change significantly. The final essay uses the before and after opening panel data to examine socio-psychological aspects of travel behavior change in response to the Expo Line opening. Random effects models of transit use, car driver trips, and active travel trips all show that the socio-psychological constructs hypothesized in the PIA model do have a significant impact on travel behavior. In addition, cross-lagged models designed to examine the attitude-behavior relationship show an apparent causal pathway from attitudes to behavior for all three travel outcomes.