Building a Visual Neighborhood Disadvantage Index to Understand Educational Disparities in North Carolina

Educational inequality is deeply intertwined with neighborhood conditions, yet widely used measures of neighborhood disadvantage, such as the Area Deprivation Index (ADI), rely on infrequently updated survey data and fail to capture the physical environments that children encounter daily. Sociologist Emma Zang will develop and validate a visual, AI-powered index of neighborhood disadvantage at the census block group level from 2007 to 2024 that will integrate Google Street View (GSV) imagery, satellite-based environmental data, and socioeconomic indicators from the American Community Survey (ACS). This multidimensional approach captures both street-level signs of disinvestment (e.g., poor building conditions, vandalism, litter) and broader structural and environmental features (e.g., lack of green space) and combines them with structural indicators of economic and institutional resources. Zang asks: How much can visual and environmental data enhance the measurement of neighborhood disadvantage relative to traditional survey-based indices like the ADI or poverty rate? Do features of visible disinvestment and ecological stress explain additional variation in students’ academic achievement and behavioral outcomes beyond SES-based measures? Which dimensions of neighborhood disadvantage most strongly predict different types of student outcomes and do these associations vary by age or school-level socioeconomic context?