PURPOSE: To examine the association of personal values, the built environment, and auto availability with walking for transportation. SETTING: Participants were drawn from 11 U.S. metropolitan areas with good transit services. SUBJECTS: 865 adults who had recently made or were contemplating making a residential move. MEASURES: Respondents reported if walking was their primary mode for nine trip purposes. "Personal values" reflected ratings of 15 variables assessing attitudes about urban and environmental attributes, with high reliability (ot = 0.85). Neighborhood form was indicated by a three-item scale. Three binary variables were created to reflect (1) personal values, (2) neighborhood form, and (3) auto availability. DESIGN: The association with walking was reported for each of the three variables, each combination of two variables, and the combination of three variables. An analysis of covariance was applied, and a hierarchic linear regression model was developed. RESULTS: All three variables were associated with walking, and all three variables interacted. The standardized coefficients were 0.23for neighborhood form, 0.21 for autos per person, and 0.18 for personal values. CONCLUSION: Positive attitudes about urban attributes, living in a supportive neighborhood, and low automobile availability significantly predicted more walking for transportation. A framework for further research is proposed in which a factor representing the role of the automobile is examined explicitly in addition to personal values and urban form.
PURPOSE: To examine the association of personal values, the built environment, and auto availability with walking for transportation. SETTING:Participants were drawn from 11 U.S. metropolitan areas with good transit services. SUBJECTS: 865 adults who had recently made or were contemplating making a residential move. MEASURES: Respondents reported if walking was their primary mode for nine trip purposes. "Personal values" reflected ratings of 15 variables assessing attitudes about urban and environmental attributes, with high reliability (ot = 0.85). Neighborhood form was indicated by a three-item scale. Three binary variables were created to reflect (1) personal values, (2) neighborhood form, and (3) auto availability. DESIGN: The association with walking was reported for each of the three variables, each combination of two variables, and the combination of three variables. An analysis of covariance was applied, and a hierarchic linear regression model was developed. RESULTS: All three variables were associated with walking, and all three variables interacted. The standardized coefficients were 0.23for neighborhood form, 0.21 for autos per person, and 0.18 for personal values. CONCLUSION: Positive attitudes about urban attributes, living in a supportive neighborhood, and low automobile availability significantly predicted more walking for transportation. A framework for further research is proposed in which a factor representing the role of the automobile is examined explicitly in addition to personal values and urban form.