Hot Spot Analysis for SNAP Nutritional Allocation
How can we ensure that food assistance effectively reaches the populations that need it most? Spatial statistical analysis offers an objective answer by identifying significant clusters of program participation, enabling evidence-based resource allocation.
In this project, my objective was to identify statistically significant spatial patterns of SNAP (Supplemental Nutrition Assistance Program) participation at the US county level, to help decision makers distribute resources more efficiently and equitably.
To achieve this, I used two complementary pattern detection methods in ArcGIS Pro. Hot Spot Analysis (Getis-Ord Gi* statistic) identifies spatial clusters of high values (hot spots) and low values (cold spots). Outlier Analysis (Anselin Local Moran's I statistic) detects clusters as well as spatial outliers — features with values significantly different from their neighborhood. I tested different distance bands (200 km default, then 100 km) to capture patterns at different spatial scales.
The results reveal distinct patterns: the southeastern United States shows statistically significant clusters of high SNAP participation (hot spots in red), while the north-central region displays clusters of low participation (cold spots in blue). The outlier analysis also identified atypical counties deserving particular attention.
Technologies Used:
SHARE
Create applications and maps to tell the story of data and transform it into action levers