The Importance of Statistical Design in Spatial Analysis: Indiana Superfund Site Locations and Socioeconomic Variables
Primary Submission Contact
Faculty Sponsor Email Address
Arts and Sciences
Geography and Environmental Science
Environmental justice research has focused on the question of whether marginalized people are more likely to live close to polluted sites. These spatial disparities have been found when analyzing the distribution of Superfund sites with race and socioeconomic status. However, these results appear to be dependent on what methods and units of analysis are used. This study examines whether these disparities exist in Indiana by comparing the results of two different statistical tests: a t-test and bootstrapping analysis.
Using t-tests to compare median income in areas close to and far from Superfund sites at distances of 1 and 5 miles yielded highly statistically significant results, indicating that proximity to Superfund sites is a significant predictor of lower median income in Indiana. Refining our initial approach shored up issues related to the independence assumption through bootstrapping and looked at additional variables. However, no statistically significant differences were observed using these methods.
This project highlights the importance of appropriate statistical design when examining relationships between geographic and socio-political variables. Further work will take a multifactorial approach and will compare values around Superfund sites to the distribution of values for a simulated set of random sites.
Johnston, Jenna and Hall, Cecilia, "The Importance of Statistical Design in Spatial Analysis: Indiana Superfund Site Locations and Socioeconomic Variables" (2019). Fall Interdisciplinary Research Symposium. 117.
Additional Presentation Information