Built environment correlates of active school transportation: neighborhood and the modifiable areal unit problem

Researchers, practitioners and community-based organizations have emphasized built environment interventions to encourage active school transportation, a practice that can contribute to the overall physical activity needs of children. This paper examines the potential influence of the modifiable areal unit problem (MAUP) on statistical modeling of the relationship between the built environment and walking/cycling to school. Binomial logistic regressions of school travel mode choice for children aged 11–12 years, in the City of Toronto, Canada, were estimated, using six spatial units for measuring built environment characteristics. The results were suggestive of the presence of MAUP across different geographical units. Travel distance, block density, signalized intersections, walking density, and schools in low-income neighborhoods were associated with active travel. This research improves understanding of the sensitivity of school travel behavior analyses to the spatial representation of the neighborhood construct.

► Children’s active school transportation (AST) behavior was examined.
► The built environment was measured using six different scales and two zoning systems.
► The MAUP may exist in statistical analysis of the built environment–AST relationship.
► Distance and several neighborhood features were associated with active mode choice.