Teun Remmers

108 | Chapter 7 children from the same school share large parts of their physical environment (Figure 1). Therefore, these shared school-environments provide unique opportunities for investigating relationships between ASPA and the physical environment. Location of schools and respondent’s residences were geocoded, and we computed a 400, 800 and 1600 meters crow-fly buffer around each school using ArcGIS (ESRI ArcGis Desktop 10.2. Redlands, CA). We subsequently classified each residence to be located 1) inside the 400 meters buffer-area, 2) outside 400, but inside 800 meters, 3) outside 800, but inside 1600 meters, 4) outside the 1600 meters buffer area. As crow-fly distances may be misleading because of barriers in the environment (e.g. highways or canals), we also computed network-distance as the shortest network distance in meters via the street network from each child’s residence to their school using Google Maps (GoogleMaps, 2015), and recoded distances in four categories. In order to keep sample sizes within categories comparable with the crow-fly distance, we based categorization on equal frequency distributions. Figure 1: Geographical location of included schools and participants