Background
Home-to-school distance is consistently identified as the strongest barrier to active school transportation. However, what is a walkable distance remains unclear as most studies used subjective measures of distance. It is possible that this threshold may vary by physical and socioeconomic contexts. A better understanding of the threshold of walkable distance will inform school planning and development. In addition to distance, measures of other environmental factors have also relied on subjective measures or objective yet aggregated measures (i.e., not specifying the environmental measure for individual students’ school travel).
Objectives
1) To identify the threshold of walkable distance for elementary school children and examine if this threshold varies by physical and socioeconomic contexts; 2) to examine the impacts of individualized, objective school travel environment and personal and social factors on children’s walking-to-school behaviors.
Methods
A combination of parental survey and GIS analyses (n=2,597) was used in this cross-sectional study. The survey was conducted in 20 public elementary schools in the Austin Independent School District. The school sample is representative of the diverse physical and socio-demographic characteristics of the district. The survey collected information for students’ school travel modes, personal and social factors, and parents’ perceptions of school travel environment. Based on the address information from the survey, students’ home-to-school routes were generated in GIS and used as individualized units for the GIS analyses of objective school travel environment. The GIS-generated shortest routes showed moderate to high validity when being compared with the actual home-to-school routes drawn by parents (n=30).
Bivariate and descriptive analyses were used to examine the threshold of walkable distance and the bivariate associations among the objective distance, the parental perception of walkable (close enough) distance, and the outcome of walking to/from school. Structural Equation Modeling (SEM) was conducted to examine the impact of individualized, objective school travel environment, and personal and social factors on walking to/from school. Clustering effect by school was addressed using the multilevel SEM approach.
Results
The percentages of “Yes” and “No” for “parental perception of close-enough distance” and “behavioral outcome of walking to/from school” in each distance range (0.125 mile interval) helped to visualize the threshold of walkable distance (Figure 1 and Table 1). The threshold is defined as the point where the percentage of “No” for “perceiving distance as close enough for walking” (or “actually walking to/from school”) starts to exceed the percentage of “Yes” (Figure 1).
The walkable threshold for the “outcome of walking to/from school” is 0.5 mile for the whole sample, and is 0.625 mile and 0.375 mile respectively, for the inner-city schools and the suburban schools. Compared to inner-city schools, suburban schools have a much higher threshold for “parental perception of close-enough distance” (1.125 mile for suburban vs. 0.625 mile for inner-city), likely because other environmental features (safety, aesthetics, maintenance) are more favorable in suburban neighborhoods. However, the threshold for “actually walking to/from school” is higher in inner-city schools (O.625 mile) than in inner-city schools (0.325 miles), possibly because some lower-income children in these schools are “captive walkers” with no transportation options except walking.
Figure 2 presents SEM results. Among individualized, objective environmental measures, the objective distance (OR=0.375, P<0.001) and the presence of freeway barrier (OR=0.394, P<0.5) are two significant barriers to walking to/from school. Among personal and social factors, parental education, car ownership, personal barriers (time constraint, etc.), and school bus service availability are negative correlates, while the number of family members, positive attitude toward walking, and positive peer influence are positive correlates.
Conclusions
The analysis of individualized, objective physical environment re-confirmed the negative impacts of distance and freeway barrier on walking to/from school. The results suggest the importance/potential of centrally-located neighborhood schools and barrier (freeway)-free attendance areas in promoting active school transportation. The context-specific thresholds for walkable distance provide important evidence for future school planning and development.
Support/Funding
This study is supported by a grant from the Robert Wood Johnson Foundation Active Living Research Program.
