Educational policy has increasingly focused on holding teachers and schools accountable for student performance. Yet popular and academic writers have long connected exposure to neighborhood violence to poor student performance. Newly available datasets, statistical methods and computer technology allow for greater power and additional control in analyzing this relation. Using school and neighborhood data (N = 792,374 students from 1,240 school neighborhoods) from New York City between 2006 and 2010, multilevel models were used to test whether exposure to violence in the school neighborhood (the number of police-reported felony assaults, homicides, rapes and robberies) predicts student performance (scaled scores on annual English and math tests). Violent crime is significantly associated with negative students outcomes controlling for a host of student and school neighborhood level variables including poverty and prior violent crime. Effect sizes were larger when predicting math outcomes than English, and for students in middle school as opposed to elementary school. These findings suggest that educational policymakers must distinguish exposure to violence from teacher and school effects and that neighborhood violence must be addressed by stakeholders of child development whether in schools or in society at large.
| Identifer | oai:union.ndltd.org:columbia.edu/oai:academiccommons.columbia.edu:10.7916/D8BV7PPN |
| Date | January 2013 |
| Creators | Chen, Jondou |
| Source Sets | Columbia University |
| Language | English |
| Detected Language | English |
| Type | Theses |
Page generated in 0.0017 seconds