Without context, your numbers have little value. I can tell you that 500 people in my county have been affected with COVID – but is that a lot? How many people are in my county? How does that compare to other areas in the US? How about to other viruses?
A few weeks into the pandemic, Tableau community members put together the Coronavirus (COVID-19) Data Hub. This data hub has COVID case data cleaned and ready for connecting in Tableau. Of course, I had to take my own pass at looking at the data, and I came up with the dashboard below.
My Tableau dashboard only goes up to a few days in March. Shortly after I created my dashboard, county-level data was no longer being updated on a daily basis, so I paused my dashboard’s data refreshes.
Mistakes I Made Building My Dashboard
As soon as I saw that COVID data was available to Tableau, I connected and immediately started building random bar charts and tables to understand the data I was working with. The first mistake I made was ignoring any data documentation, and just diving in.
About five minutes into my exploration, I turned to my husband and said, “Wow, did you know that we had 2000 cases in our county last week?” At the time, this was a lot higher than the numbers that were being reported, and I was surprised at what I had found.
It turned out, the dataset was keeping track of 4 types of cases: Active, Deaths, Recovered, and Total. Total was a row combining all case types – I was doubling my values. I was already spreading false information, and it was due to me not fully understanding the data set before sharing my findings.
Recognizing how easy it was to make mistakes with the data, it made me hesitant to trust other novice COVID visualizations. This reminded me that not only do we need to make sure we trust the source of the data, but trust the designer of the data as well.
Share Data Responsibly
After creating my dashboard, I also came across the article 10 considerations before you create another chart about COVID-19. It is a great read regarding using data around a sensitive topic. While it’s a timely data set to work with, you want to be considerate of the fact that you don’t want to encourage panic and that you don’t want to mislead people. I highly recommend checking it out if you’re also interested in working with this data.
