Abstract

From exploratory spatial data analyses and geographically weighted regression (GWR), we found that previously hypothesized relationships between socioeconomic status (SES), race, urbanization and mortality were present and significant in the Atlanta metropolitan area for 1995–1999 and that the relationships between these predictors and mortality varied spatially, such that distinctive geographic patterns emerged. These patterns reflect the spatial processes operating in Atlanta for the past few decades, namely, rapid residential and commercial development in the outer portions of the metropolitan area and a concurrent movement of the affluent white population away from the central city, leaving behind a predominantly African American population with low SES. We also found that the relative influence of each predictor on mortality varied spatially, with SES demonstrating the most dominant influence in the majority of the study area and race demonstrating the most dominant influence in and near the City of Atlanta.

Keywords: Atlanta; Urbanization; Mortality; GIS; Geographically weighted regression

Article Outline

1. Introduction

2. Methods

2.1. Data sources

2.2. Standardized mortality ratios

2.3. Selection of independent variables

2.4. Urbanization variable

2.5. Socioeconomic status (SES) index

3. Analysis

3.1. Spatial autocorrelation

3.2. Geographically weighted regression

4. Results

5. Discussion

6. Conclusions

Acknowledgements

References