Title:
What statistical and spatial relationships exist between health insurance, race, income, and education in the state of Georgia immediately before and after the implementation of the Affordable Care Act?
What statistical and spatial relationships exist between health insurance, race, income, and education in the state of Georgia immediately before and after the implementation of the Affordable Care Act?
Author(s)
Walker, Evan
Advisor(s)
Editor(s)
Collections
Supplementary to
Permanent Link
Abstract
What statistical and spatial relationships exist between health insurance, race, income, and education in the state of Georgia immediately before and after the implementation of the Affordable Care Act?”. To answer this question, two datasets were used. They were both five-year estimates from the American Community sSurvey. The first range was for 2009-2013, and the second was an estimate from 2012-2016. The data obtained was for the 1959 census tracts in the state of Georgia. These years were chosen because the ACA was implemented in 2014, therefore the first dataset would not be affected by the ACA and the second would what largely be after its implementation. This study combined both linear statistical analysis as well as spatial statistical analysis. The variables chosen were income, race, education level, and health insurance. More specifically: average income for each tract, percent non-white/minority population, percent of individuals over 25 years-old with less than a high school diploma or GED equivalent, and the percentage of the population that in uninsured. These were chosen because I felt that they are all suitable metrics for examining these complex socio-economic factors. In the linear regression analysis health insurance was the dependent variable (DV) in all the regressions. For each dataset several combinations of the independent variables (IV) were used, in addition the difference between variables in the two time periods was regressed, and finally a logistic regression was performed on the differences between the two time periods. Unfortunately, the regression produced very little correlation amongst any of the variables. (This will be discussed more thoroughly in the results section). The next part of the analysis was the spatial analysis for each variable a get-is Ord hotspot analysis was performed, a Moran’s I test for spatial autocorrelation, and then individual choropleths were generated for each variable as well.
Sponsor
Date Issued
2018-08
Extent
Resource Type
Text
Resource Subtype
Masters Project
Capstone Paper
Capstone Paper