Once I developed the two models in ArcGIS I described in my two previous posts, I was able to analyze the relationship between health care accessibility and development assistance for health in Malawi. I used the statistical toolkit included in ArcGIS to run analysis on the two variables. I wanted to determine if development assistance for health was flowing to areas where preexisting health care facilities are inaccessible to large populations.
-
- To begin, I tested for spatial autocorrelation of development assistance for health by district in Malawi. The results show that the amounts are distributed randomly, meaning that the values for each districts are not heavily influenced by the values of the districts nearest to them because they are close in space.
-
- I ran Ordinary Least Squares Analysis to model the relationship between accessibility and aid. This map shows how well each data point (each district) matches to the model created by OLS analysis. It is evident that In a few regions, especially near the large cities in the south, the predicted values do not match well with the observed values.
-
- These figures produced during OLS analysis clearly show that there is no obvious relationship between health care accessibility and that more data is needed. For these reasons, I am omitting more involved statistical analysis.
