Using GPS-enabled Activity-Travel Surveys to Understand Individuals’ Spatiotemporal Behaviors and Promote Equitable Mobility

Abstract

The GPS-enabled activity-travel surveys have emerged as new trending data to understand travel patterns at the individual level. However, it also introduces new challenges in data quality control, data analytics, and data interpretation. This dissertation addresses these gaps through three interconnected projects aimed at enhancing our understanding of individuals' spatiotemporal behaviors and promoting equitable mobility. First, since the smartphone-based survey integrates auto-detected trips and activities with user-entered contextual information, new data quality issues arise and need to be dealt with after data has been collected and before has been used for analysis. By recognizing the lack of systematic approaches to handle data quality issues in GPS-enabled surveys, this dissertation develops a framework to handle these quality issues to ensure attribute completeness and logical consistency. The framework applies statistical and data mining methods to detect quality issues instead of arbitrarily setting threshold values. Second, the dissertation challenges traditional assumptions in time geography by proposing a method to derive personalized space-time prism anchors. By utilizing the spatial, temporal, and thematic information in the activity-travel data, this approach reveals complex patterns of routine behaviors that existing models often fail to capture. Finally, this dissertation examines gender differences in travel behaviors through an intersectional lens, employing innovative methods such as sequence alignment and Chi-square automatic interaction detection. The analysis uncovers gendered mobility patterns that go beyond simple trip-based comparisons, addressing a critical gap in transportation equity research. In sum, this dissertation advances both methodological approaches and theoretical understandings in urban mobility research. By improving data quality and refining conceptual and methodological frameworks, this dissertation contributes to the development of more equitable, efficient, and responsive urban transportation systems. The findings have significant implications for transportation planning and policy, offering tools and insights for creating smarter, more inclusive cities.