When I was an undergraduate a number of years ago, I took a statistics course. On the first day of class, my professor said something that has stuck with me.
“Statistics,” he asserted, “can be used to create any impression desired. That means they can also be used to manipulate facts and hide truth.”
I’d registered for the class only because my major required it, but this guy now had me riveted.
“Even if you don’t end up using what you learn in this class on a professional basis,” he warned, “you’d better know how to read statistics as a life skill—to be able to discern how other people leverage data. Otherwise, you’ll always be vulnerable to deception.”
To clarify, one is always called to make decisions about how to analyze and present data. A multitude of choices often exist. However, to my former professor’s point, one can choose responsibly or irresponsibly. And when you start looking around, you do tend to notice that statistical data gets abused with frequency. Hence Mark Twain’s famous assertion: “There are three kinds of lies: lies, damned lies, and statistics.”
This post is a first effort to help people take a closer look at Public Health of Madison and Dane County’s (PHMDC’s) public-facing COVID dashboard. It’s important to understand what we’re actually seeing and not seeing on the dashboard, because for nearly two years PHMDC has used the dashboard as a primary tool for broadcasting the virus’ impact within this community. Local media outlets and others rely on the dashboard PHMDC has provided. With even a bit of knowledge, it becomes possible to see that the picture PHMDC paints for you via the dashboard is—well—problematic and more than a little deceptive.
In case you haven’t been following the dashboard for these past many months or need a refresher in how it looked early on, here’s a sample screenshot of the front page as it appeared on May 22nd, 2020.
Here’s a second screenshot from June 14th, 2020. It provides the same basic information array, with the exception that the bar graph along the x-axis has been dropped.
Take a close look at the above screenshots. Notice where your eye first travels? Almost certainly, it’s to the ever-rising line of cumulative cases at center screen. It goes up and up and up every day…right? Scary! From there, your eye likely travels to the large numbers in great big type on the left side of the screen…953 cases, 175 hospitalized, 30 deaths. YIKES! Even that huge testing number is meant to send a signal. We wouldn’t have to do all that testing unless there was a big problem…right?
What is the picture these items paint when taken together? An exploding pandemic. And what reaction does perception of an exploding pandemic naturally trigger? Anxiety and fear would be pretty understandable reactions, yes? PHMDC is painting a picture of how bad the virus already is within Dane County, and they clearly want you considering how much worse it could get.
In case you think that this data presentation is random or accidental, disabuse yourself of that notion. PHMDC has at least two professional epidemiologists on its data team, one of whom, Katarina Grande, is the team lead. Epidemiologists possess specific expertise in crunching and presenting data. It’s very clear, in this case, that PHMDC wants you to connect these two key items—the big, scary numbers and the perpetually rising graph line. You’re meant to understand that your life and health are at risk—and not just a small risk, either.
I could say a lot about this first early dashboard screen, but I’ll focus on a single truth that packs a wallop: Numbers without context are just numbers. They’re not big or scary at all. Context alone gives numbers power in our hearts and minds.
In this case, the context PHMDC has provided for the numbers on the left of the screen is that terrifying cumulative case line that just keeps going up and up and up.
However, we need to take two key realities into consideration.
First, the only direction a cumulative line graph ever goes is up; they’re largely useless. They’re kind of like telling your viewer that water is wet; when you add things together, they increase. Presenting a cumulative case graph as the context for understanding the impact of COVID-19 in Dane County amounted to a cheap but effective parlor trick. Prominently presenting an essentially worthless graph created a desired impression for Dane County residents that the virus was a greater threat than it actually was on any given day. In fact, this illusion worked so well for PHMDC that the agency continued to leverage it until January of 2021, at which point it finally revamped the dashboard and moved on to related devices.
Second, to provide immediate, responsible context for case, testing, hospitalization, and death counts, PHMDC needed to tell you the total population of Dane County. Doing so would have facilitated simple calculation of percentages, thereby revealing pretty much outright just how low the risks actually were. Those risks, incidentally, have remained quite low for every single Dane County demographic throughout the course of the pandemic—a topic I’ll tackle in an upcoming post.
Until then, Dane Undivided wishes you a Happy Thanksgiving. We hope you’ll share this article with others and talk about it around the table today and all weekend long. In fact, we hope it will inspire you to take a close look at PHMDC’s COVID dashboard all on your own and see what other things you notice.
On this day, as we all count our blessings, we also encourage you to re-embrace the gift of your inherent freedom. Refuse to buy the fear and control PHMDC has been pushing.
Stop complying. Start defying.
Encourage others to do the same.
More coming. Stay tuned…
Comments