Assessing School Communities Using Google Street View: A Virtual Systematic Social Observation Approach

Abstract

Little research in education has focused on school neighborhoods. We employ a novel systematic social observation tool—the internet-based School Neighborhood Assessment Protocol (iSNAP)—within Google Street View to quantify the physical characteristics of 291 preschool communities in nine U.S. cities. We find low to moderate correlations (r = −.03 to −.57) between iSNAP subscales and Census tract poverty, density, and crime, suggesting that the characteristics captured by the iSNAP are related to, yet ultimately distinct from, existing neighborhood structural measures. We find few positive associations between iSNAP community characteristics and 1,230 low-income preschoolers’ end-of-year outcomes. Specifically, resources for outdoor play (e.g., playgrounds, open fields) on school grounds predicted stronger child self-regulation skills, whereas global ratings of safety and care for both the school grounds and surrounding neighborhood predicted stronger approaches to learning skills. Indicators of physical order were not associated with child outcomes.