How to process all these data and statistics about COVID-19, for the non-scientist

We are being utterly bombarded on a daily basis with information about the pandemic. It’s one thing to talk about it for fun (because it is so fun to talk about COVID), but another to use it to make decisions.

I was inspired by a conversation I had with my grandma recently about not letting fear get the best of us, and I totally agree with her. So, at the root of this post is using the information versus letting the information get your undies in a bundle. That’s a choice we have to make each day to stay sane, I believe! I’m also the kind of person that takes comfort in making informed decisions in general.

So, here we go. Let’s talk statistics.

On statistics: there are probably about a million different ways we can look at the COVID data we have available and summarize it to make something meaningful. But that is just it. Which statistics are the most meaningful, and which are just noise? A year ago, I might have said I’m not sure I can help you there without recommending 3 or 4 statistics courses, but I’ve been doing some thinking and maybe that’s not actually necessary. You have an incredible tool at your disposal right this moment: critical thinking. I truly do believe we all have this ability, but we have to practice it to get good at it.

Something I learned while in grad school was that when I’m looking at a dataset, or other people's statistics for that matter, I can get lost very fast if I don’t have a plan. The way I learned to cope with this was to skim through the data to see what was there, and then take a good rest from it completely and later, when I was not looking at the data, draw a graph of what I expected to see. I learned this on my own through trial and error; I don’t remember it being something anyone taught me outright. Drawing the graph helped me see which statistics would be most meaningful in answering the question. So why not try this yourself?

Think of a practical question you have about COVID. Maybe it is whether you should go into the grocery store or order grocery pickup (which would entail you coming into contact with one person instead of upwards of 50, depending on the day and time you go and which store you’re going to, of course). To me, this is a meaningful question because I would prefer to go into the store so I can pick out the produce myself and use my own bags at check-out, but not if I think I might catch the virus (or give it to another person unknowingly) in doing so. So, there’s a trade-off I’m making by choosing pickup. Some stores have fees for pickup (mine doesn’t), but that could be a trade-off too. Okay, so, what is the likelihood that any of those 50 people will have COVID versus the one person bringing out your groceries? Personally, the way I would answer that question is by figuring out the number of people that likely have COVID right now out of 50. That means I need to know how many estimated active cases there are out of the total number of people, calculate the percentage, and extrapolate to 50 people (big words, but not difficult! You can handle this math!). The closest we can probably get at this point to understanding the situation for each of our own little communities would be county-level, as state-level would probably be too large (especially when you live in Texas, like me) and I haven’t seen any numbers at the city-level, although each state probably varies a bit in what they report, so think critically about how you can use what you have!
​
So this is where I got a little lost. I found a snapshot of the Texas DSHS dashboard and started trying to make sense of it, and pretty soon I was forgetting what I was trying to do. As soon as I realized my feelings of being overwhelmed, I zoomed back out, took a breath, and reassessed. I did note that the data might come in handy later, though!

I found a “trends by county” page for Texas DSHS which is totally awesome and I hope your state has one too. If not, write to your government officials! I found a “total cases per 1000 people” value by county which happens to be exactly what I’m looking for. Didn’t think I would get that lucky. As of July 17th (yesterday), there were about 14 cases per 1000 in my county. So that’s about 1.4 people out of 100, or 0.7 people out of 50. I would round that up to 1 person. Statistically speaking, chances are that if you go into a place that has 50 people (in my county), one person will have it, and the one person who brings out my groceries for pickup probably won’t (although, maybe you could say that if you go to the grocery store 50 times, one of those times the person bringing out your groceries will have COVID? But that’s what I call getting lost in the statistics! Just wear a freaking mask).

If you're nervous every time you pass someone on the sidewalk, consider that there is a 1/71 (14/1000) chance that the person you just passed has the virus, that is, if your county's number look like mine. Those odds can mean something different to each of us, since we each have different thresholds for the risks we want to take.

What you choose to do with this information is your decision. But isn’t it kind of nice to know that there’s a way to use data to make your decisions, instead of just worrying all the time?

Importantly, we could go on and on about why these numbers might not be accurate, and I believe that is also worth discussing. But if it’s not going to help answer a practical question then it defeats the purpose of this blog post. I guess, to a degree, I am choosing to rely on my faith in the authorities to put out numbers that are at least close enough to help me make this decision. Anyway, my intuition tells me it’s probably close enough!

So, I know this is a little oversimplified, and some of our questions are more complex than that and we kind of need to rely on experts to answer them for us. Here’s another trick you can use: the hierarchy of evidence. Too much to write here, so that’ll be one of my next posts.
​
Have questions or criticisms about these calculations? I would love to hear your thoughts. Comment below! Let’s discuss!