I have combined the following data into one data matrix:
– 2009 Vaccination data table (subset of most often given vaccines reflecting the trend) ;
– American Human Development Index by State from American Human development Project;
– Classification of Blue and Red states;
Second, Principal Component Analysis was applied to all data (vaccination rates and human development index) with autoscaling.
PC1 captures 39% of variance in the data and separates samples by those having high rank (cumulative index), high Education, Income and Health index from those having low Rank. Mostly blue states and some red states (AK, ND, NE, UT, KS) have highest rank. Vaccination rates do not contribute into PC1 (close to 0) indicating that there is no direct correlation between human development index and vaccination.
PC 2 separates states by vaccination rates. Those on top have higher vaccination rates that those on the bottom of a biplot. There is week correlation (captured in ~16% of variance in the data) between vaccination rates and education and income index and rank.
4 groups of states are classified by PCA:
1. Red states having quite good vaccination rates and very bed HD index.
2. Mostly blue states and some red states having best vaccination rates and best HD index.
3. Purple states having worst vaccination rates and worst HD index.
4. Mixed states – some blue, some red and Co with very bed vaccination rates but high Health Index.