Social Sciences · Transportation
Determinants of Active Travel to School
Dianne S. Ward, Amber Vaughn, Brian Neelon, Laura Linnan, Janet E. Fulton, Sarah Martin
See on the mapAbstract
BACKGROUND: The proportion of students walking and bicycling to/from school has declined substantially over the past 30 years. Distance and parents' fear of crime and traffic danger are often cited as barriers. Counter measures to reduce these barriers and increase active travel (AT) to school are needed. In order to create effective interventions targeting AT, a better understanding of the determinants of AT are needed. PURPOSE: To compare the characteristics of students who use AT and their parents to those who do not (NAT). METHODS: Schools (13) representing 5 geographic regions of the US were randomly chosen from a convenience sample of respondents to a Walk to School (WTS) survey. WTS coordinators provided school level information and parents of 4th and 5th grade students completed a brief survey. School level data (size, % free or reduced lunch, walkability assessments, WTS Day participation, etc.) and parent level data (child gender and age, distance between home and school, perceived barriers of AT, perceived norms, etc.) were compared based on parent-reported AT status. After assessing differences between AT & NAT, we used a Generalized Estimating Equation (GEE) procedure to identify significant predictors of AT. To build the model, we used a split-sample validation approach. RESULTS: Five significant covariates for AT were included in the final model: gender, distance, perceived barriers around personal safety, and two norm-related variables-the perceived use of walking or bicycling for school travel by other children in the neighborhood and the reported frequency with which neighbors walked or bicycled with kids to school. Boys were twice as likely to use AT compared to girls (OR =2.1, CI: 1.29, 3.42). The odds of AT increased by a factor of 2.73 (CI: 1.95, 3.86) for every 1/2 mile closer the child lived to school. Those who identified fewer personal safety barriers were significantly more likely to use AT (OR = 1.18, CI: 1.05, 1.32). The odds of using AT increased three-fold when parents noted the presence of other children walking or bicycling in the neighborhood (OR = 2.88, CI: 2.12, 3.91). Finally, those who reported higher frequency of neighbors walking or bicycling with children to school had higher odds of AT (OR 1.34, CI: 1.14, 1.58). CONCLUSION: In addition to the association of gender and the previously described barriers of distance and safety, perceived norms about walking/bicycling to school are important predictors of AT. These findings suggest that efforts to increase AT may need to address personal as well as community factors. Funded by the Centers for Disease Control and Prevention, SIP 9