عنوان مقاله [English]
Analyzing the casual relationship between built environment and travel behavior has been a key issue in the transportation and planning related literature. Residential self-selection is one of those issues that needed more elaboration in built environment and travel behavior studies. This concept means people choose their neighborhoods according to their travel attitudes and preferences. In other words, people who prefer to walk more, decide to reside in walkable neighborhoods. Therefore, people walk more because they like to walk. Hence, residential self-selection, which refers to the propensity of people to choose where to live is based on their travel attitude. So finding the influences of built environment on travel behavior while taking into account the role of travel attitudes, neighborhood preferences and socio-demographics characteristics in explaining this relationship is vital. In this paper, to understand the extent to which the observed patterns of travel behavior can be attributed to the residential built environment, structural equations modeling is applied to investigate the relationship between the built environment and non-work travel frequencies by walk mode in three neighborhoods with different land development patterns in Tehran Metropolis. By identifying the direct and indirect impacts of dominant factors on travel behavior, the following assumptions are presented: 1) if a built environment element affects the travel behavior or other factors directly or independently, then that element has a causal relationship with travel behavior, 2) if travel attitudes or neighborhood preferences impact on build environment, then it could be understood that self-selection is confounding the casual relationship between built environment and travel behavior. To assess these two assumptions, the data compiled from 273 inquiries in three neighborhoods (Monirye (as a traditional neighborhood), Golestan (as a car-oriented neighborhood) and Bime (as a conventional neighborhood)). By using exploratory factor analysis, the built environment dimensions are extracted as follows: residential environment characteristics, highway accessibility, transit accessibility, destination diversity and their accessibility, denseness and neighborhood preferences domains. The latter one refers to what are the accessibility preferences/priorities of residents if they want to move to a new neighborhood. Besides, the travel attitudes factors elicited by exploratory factor analysis are found to be: pro-walk/bike/transit, car dependent, travel minimizing, pro-automotive mode. After determining the domains of built environment, accessibility preferences/priorities and travel attitudes, structural equation modeling is applied to understand the relative and casual relationship between built environment and travel behavior in 3 neighborhoods. Although the evidences from car oriented and conventional neighborhoods show the causality of relationship between built environment and travel behavior, in traditional neighborhood travel attitudes and neighborhood preferences influence on travel behavior directly and indirectly. Nevertheless, the whole comparison assessment of direct/indirect impacts on travel behavior in the three studied neighborhoods shows that built environment’s elements have casual effects on travel behavior. As an instant, destination diversity and their accessibility has direct and indirect impact on travel behavior, respectively, in traditional neighborhood and two latter types. Taken together enhancing the diversity, transit accessibility, and reducing highway accessibility play a more prominent role in non-motorized travel behavior. In general, our results suggest that if cities use land use policies to offer more options to use non-motorized modes, many residents would tend to do so.