Geographic Information System Methodologies and Spatial Analysis in Health and Environmental Disparity

Studies reported that racial/ethnic minorities living in disadvantaged neighborhoods experienced a greater rate of exposure to environmental hazards. Knowledge of environmental exposure risks, distributional patterns and their effects on population health require a geographic perspective while investigating social injustices to better understand the causes of health disparities among different populations. However, previous studies often fail to recognize processes and assumptions of spatial analyses. In this paper, we demonstrated the importance of such processes. We used exploratory spatial data analysis methods to examine potential spatial patterns of demographic and cancer risk distributions in Chicago. First, we examined the presence of overall spatial clustering using Moran’s I statistic. Our Global Moran’s I statistic showed clustering for percent poverty, percent black and non-point cancer risk in predominantly poor neighborhoods in Chicago. Local autocorrelation was conducted to identify spatial clusters and spatial outliers. Local indicators of spatial association provided univariate significant maps, cluster maps and scatterplots which identified spatial clusters for percent poverty, percent black and non-point cancer risk in Chicago. We then conducted bivariate analysis which showed that standardized high percent poverty was significantly correlated with a standardized high neighboring non-point source cancer risk. These findings were conclusive evidence that indicated the presence of spatial clusters, while the strengths of the associations cannot be determined. The findings warrant further analysis with spatial regression methods.