On Christmas Eve, I posted on the new Census Bureau poverty estimates for every county in the U.S. In the graph above, I correlated median household income with COVID Case Mortality for the 10 county region. It shows that median income predicts COVID case mortality accounting for 62.9% of the variability. County median income but not poverty predict COVID mortality.
The graph above shows a statistically non-significant association between the % in poverty and COVID case mortality in the 10 county region. This relationship only accounts for 14.3% of the variability in COVID case mortality. Comparing two of the above graphs, we see a lot more spread around the regression line with the percent in poverty than with median income. As previously reported, Centre county had the second highest poverty rate but the highest median income rate.
Vaccination Rate
Median income came close to being a significant predictor of COVID vaccination rates accounting for 31.6% of the variability. To be statistically significant, at least 38.9% of the variability would have to be accounted for. The above graph shows that Cambria and Fayette Counties have low income levels but relatively high vaccination rates. Conversely, Indiana and Bedford have low income levels and low vaccination rates. These outliers are the reason why this relationship is not significant.
As with COVID Case Mortality, the percent in poverty was a much weaker predictor of COVID vaccination rates with a much greater spread around the regression line. Westmoreland County had a low poverty rate and a high vaccination rate while Bedford County had a low poverty rate and the lowest vaccination rate. Centre County had the second highest poverty rate and the highest vaccination rate.
The purpose of this post was not only to show how poverty and median income rates predict mortality and vaccination rates. The purpose was to show what non-significant correlations look like. In the past, I reported on significant results only. A lack of significance can be just as revealing as its presence. It is to show why median income but not poverty predict COVID case mortality.