Looking from the Outside In: A Spatial Analysis of Students' Neighborhood Characteristics and School Performance in Philadelphia

Access to a quality education is unevenly mapped by the demographic context of the communities in which public school students live. A growing body of research suggests that this geographic effect deeply challenges the goal of equalizing educational opportunities. This paper explores the relationship between students' neighborhood characteristics and school performance in Philadelphia's public schools using spatial analysis methods. Data include student and school locations as well as block-level census information and Pennsylvania System of School Assessment (PSSA) scores for grades 5, 8, and 11 during 2005-2008. Multiple regression and K-function analyses in conjunction with geographic information systems (GIS) were employed to test hypotheses linking students' community attributes to their schools' academic performance. Characteristics of students' neighborhoods, such as the percentage of the population holding a 4-year college degree and the proportion of families with a female head of household, were found to be strong predictors of school performance. Across all nine datasets, the geographic diversity of the student population was positively associated with school performance. In most cases, significant clustering was detected among schools reporting scores at or below the first quartile, and the locations of the lowest- and highest-performing schools were discovered to be spatially dependent. A range of policy implications is offered. Presented at the 2010 ESRI Education User Conference in San Diego, California, July 11, 2010. The following are appended: (1) Description of the Geography and Variables; (2) Statistical Results--All Datasets; and (3) Additional Maps and Figures. (Contains 7 footnotes, 9 tables, and 31 figures.)