Concurrent validity of Neighborhood Environment Walkability Scale (NEWS) items was evaluated with objective measures of the built environment using geographic information systems (GIS).
A sample of 878 parents of children 10 to 16 years old (mean age 43.5 years, SD = 6.8, 34.8% non-White, 63.8% overweight) completed NEWS and the International Physical Activity Questionnaire. GIS was used to develop 1-mile street network buffers around participants’ residences. GIS measures of the built environment within participants’ buffers included percent of commercial and institutional land uses; number of schools and colleges, recreational facilities, parks, transit stops, and trees; land topography; and traffic congestion.
Except for trees and traffic, concordance between the NEWS and GIS measures were significant, with weak to moderate effect sizes (r = −0.09 to −0.36, all P ≤ 01). After participants were stratified by physical activity level, stronger concordance was observed among active participants for some measures. A sensitivity analysis of self-reported distance to 15 neighborhood destinations found a 20-minute (compared with 10- or 30-minute) walking threshold generally had the strongest correlations with GIS measures.
These findings provide evidence of the concurrent validity of self-reported built environment items with objective measures. Physically active adults may be more knowledgeable about their neighborhood characteristics.
Adams, Patrick, and Norman are with the Dept of Family and Preventive Medicine, University of California, San Diego. Ryan is with the Dept of Public Administration and Urban Studies, San Diego State University. Kerr and Sallis are with the Dept of Psychology, Active Living Research, San Diego State University.