Hazard-Based Analysis of Travel Distance in Urban Environments: Longitudinal Data Approach

This paper focuses on identifying important factors that determine activity-based travel distance in urban areas. Building on past research that has demonstrated the conceptual equivalence of hazard models applied to either temporal or spatial settings, the length of the distance from origin to destination is statistically modeled as longitudinal data using hazard-based modeling methods with data from Athens, Greece. Based on the data analysis, the Weibull model with gamma heterogeneity provides the best statistical fit, and a number of factors significantly affect travel distance, including socioeconomics and demographics, trip characteristics, mode choice, trip frequency, time of day for the trip, and type of activity participation. The proposed methodological approach and the research findings help to better understand travel behavior in terms of trip distance in the urban areas, an issue of significant importance for both transportation researchers and planners.