Greenness and school-wide test scores are not always positively associated - A replication of linking student performance in Massachusetts elementary schools with the 'greenness' of school surroundings using remote sensing

Recent studies find vegetation around schools correlates positively with student test scores. To test this relationship in schools with less green cover and more disadvantaged students, we replicated a leading study, using six years of NDVI-derived greenness data to predict school-level math and reading achievement in 404 Chicago public schools. A direct replication yielded highly mixed results with some significant positive relationships between greenness and academic achievement, some negative, and some null - but accompanying VIF scores in the thousands indicated untenable levels of multicollinearity. An adjusted replication corrected for multicollinearity and yielded stable results; surprisingly, all models then showed near-zero but statistically significant negative relationships between greenness and performance. In low-green, high-disadvantage schools, negative greenness-academic performance links may reflect the predominance of grass in measures of overall greenness and/or insufficient statistical controls for the moderating effect of disadvantage.