Online Program

Comparing Pycnophylactic and Kriging Methods Using Data Related to the Flint Water Crisis (304909)

Keywords: Spatial Statistics, Modifiable Areal Unit Problem

This project focuses on the Modifiable Areal Unit Problem (MAUP) using data from the Flint Water Crisis in Michigan. MAUP is a problem in spatial statistics concerned with how different types of spatial aggregation used on the same data can lead to different results. The motivation of this project follows from the Flint Water Crisis, when concerns were raised due to blood lead levels collected from people who lived within any of the ZIP codes of Flint. Since the ZIP code borders of Flint overall covered a larger area of Michigan and did not align with the city border of Flint, analyses included people who did not reside within the city of Flint and did not consume the water. Misrepresentation of the data may have produced biased, misleading results and serves as an example of how results depend on the type of geographical or spatial unit used. In this project, the pycnophylactic method which incorporates spatial smoothing functions, and the kriging method were implemented in R to transform the data from one level of spatial aggregation to another. In particular, areal polygon-to-polygon data transformation was conducted using both methods and results were compared.