Using a Neighbourhood Measure of School Readiness to Predict Academic Achievement in a Population of Children

This study investigates whether academic achievement in individual children can be predicted by the proportion of children with school readiness vulnerabilities within the school neighbourhood. The data analyzed came from: 1) an extensive administrative database on the EDI (Early Development Instrument) (Janus Offord, 2007), which provided the measure of school readiness vulnerability; and 2) a large study of academic achievement and associated factors conducted in Latin American and Caribbean countries (SERCE, UNESCO/LLECE, 2008), which provided the measure of individual child achievement and demographics at the school level. These data were collected in 2006 (SERCE) and 2008 (EDI) in Nuevo Leon, Mexico. A two-level Hierarchical Linear Modeling (HLM) approach was conducted to analyze the data, with individual SERCE student level predictors at level-1, and average EDI school neighbourhood vulnerability scores as predictors at level-2. The EDI records (n=50,400) were nested within a total of 116 school neighbourhoods comprised of 151 schools. Results indicate that the level-2 predictors (EDI) accounted for a small but significant proportion of between school variance in third (13.8%) and sixth (17.2%) grade reading. However, the level-2 predictors (EDI) did not account for significant variance in third and sixth grade mathematics. These results provide further evidence for the influence that average school or neighbourhood readiness variations have on academic achievement. The study findings add to the conceptualization of the EDI as a marker of school and neighbourhood influences and suggest an enriched perspective on trajectories in learning for the UNESCO/LLECE research initiative.Ph.D