Project Summary

As an assignment in the first semester of my MS in Business Analytics program at Boston University, we were tasked with visualizing a COVID dataset. We were free to find a dataset of our choosing, and design a dashboard, while maintaining visualization best practices. I started with a dataset containing COVID death and infection data from each state over time. In addition, I located another dataset for mask usage in each state, and a dataset containing state population data (which is used in calculating deaths as a proportion of population for each state). I merged each of these datasets together in Tableau on the state feature so that each state had population and mask usage information. Once the data was loaded, I constructed the dashboard, incorporating a variety of different visualizations.

Conclusions

I found the results of the dashboard very interesting, and in a lot of cases, unexpected. In one visualization, I displayed a map of the United States with each state color coded based on mask usage. In another visualization, I displayed a bar chart showing deaths as a proportion of population based on state. According to the data, North Dakota has low mask usage compared to the rest of the country. However, the number of deaths as a proportion of the population is very low, which is counterintuitive to the idea that high mask usage would help protect people. However, it is important to note that this analysis does not incorporate population density. New Jersey, a state with a relatively high population density has a high number of deaths as a proportion of the population. There are numerous variables that affect the deaths that occur in a state, including how many activities can be done outside given the climate. We can say that on average, with all else being held constant, locations with higher population densities see higher death rates.

Another interesting observation was in viewing the number of deaths over time for a state with high mask usage compared to one with low mask usage; these states are Delaware and North Dakota respectively. In North Dakota, we can see an exponential increase in deaths. However, in Delaware we can see that there is a large initial increase, but then the curve flattens over time.

Screenshot of the Tableau dashboard (click the button at the top of the page to view the interactive dashboard)