Interesting Questions
A couple of posts ago I wrote about why I was tracking the coronavirus using Power BI. One of my goal was to practice asking interesting questions that would bring out something unusual or astonishing or just plain interesting. In other words, try to think of something that no one else was doing because I just might learn something.
Well a few days ago, I just thought of something that I hadn’t seen done before and was curious what the effect would be. So, I’ve been creating charts like the following:
I asked, “Can I combine the two charts – the moving average of case infections and the moving average of deaths – into one chart? So here I’m changing the prior image with the addition of a new graph.
Woah! Now I can really see the time lag of deaths. Intellectually, I knew there was a time lag but to actually see it on the graph after combining the data into one graph was still startling. When I graphed the case infections and deaths separating, it was a little harder to see the time lag – maybe due to the size of the horizontal axis, but I don’t know, the axis seems to be the same. Only when I combined the data onto the same graph with the same timeline does the lag truly become apparent.
One thing about looking at these graphs: don’t pay attention to the height of the line graph, especially the deaths, because the death depiction is on a different scale than the case infection. Deaths will not be greater than case infections. The case infection scale is on the left hand side and the death’s scale is on the right hand side. The way to look at these graphs is to see the timing of the peaks and valleys. It’s the timing that can be compared.
After seeing that kind of lag, I decided to try some other states. I chose states that I thought had larger populations (California, New York, Texas, Florida), North and South Dakotas because of the whole Midwest thing, Washington because it was the first state to report the coronavirus, and Michigan because of its virulent, gun happy citizens.
Hmmm, there’s some weird results out of this exercise. The North Dakota’s graph is bonkers; I’m don’t know why the zero baseline was shifted upward. What struck me was that New York and Washington do not show that time lag as the others do. I don’t know why; that is very odd.
Anyway, that was interesting to play around.
