This study is a replication of an earlier study investigating a possible link between exposure to greenness surrounding schools and student academic achievement. The earlier study, conducted by Wu and colleagues in 2014, included data from schools in five major cities in Massachusetts. Results showed that greenness surrounding schools was positively associated with student academic achievement in both Math and English.
The current study analyzed math and reading achievement data of students in 404 Chicago public schools. Schools in the Massachusetts study had more surrounding greenness and fewer disadvantaged students than the Chicago study. Other variables, however, were similar. Both studies focused on third-grade performance in Math and English/Reading, and both studies calculated surrounding greenness using NASA's Earth Observing System's remote sensing techniques of satellite imagery that converts data into the Normalized Difference Vegetation Index (NDVI).
An analysis of the data for this study revealed no convincing evidence of a positive association between greenness and academic performance. These results were surprising, as they contradicted the findings of the earlier study. The authors suggest that perhaps the low-greenness, high-disadvantage context for this study may explain, in part, the difference in findings. They also offer other possible explanations. One possibility relates to how “surrounding greenness” is defined. NDVI measures used by these two studies do not distinguish between vegetation types; Urban Tree Canopy assessments -- used by some other studies -- do. Some research shows that trees are positively associated with academic performance but that grass and shrubs are negatively associated with performance. Chicago, where this study was located, has relatively low levels of tree cover. The surrounding greenness, then, as indicated by NDVI scores, represents more grass cover than tree canopy. Another possible explanation for the contradictory findings relates to potential moderating effects of poverty and race on the greenness-academic performance relationship. The population of the Massachusetts study represented low levels of disadvantaged students; the Chicago study included high levels of disadvantaged students.
The authors offer several specific recommendations for future research, including (1) distinguishing between tree and other forms of green cover; and (2) considering disadvantage as a moderator of how green cover affects academic performance.
Research Partner