Lies, Damn Lies, and Statistics
Saturday, September 12, 2020
I am desperate for data. This whole conversation is full of people citing studies that don’t exist, or citing studies that do exist but don’t mean what they say, or saying things that are logical but for which there’s no actual data to show.
I have been writing the number of new cases and deaths in my journal every day. Every source shows something different, but I’ve been operating on the theory that as long as I use the same source every time, at least my numbers are consistent. Consistently wrong, no doubt, but consistent.
This is particularly important to me because I want to make data-based decisions about my behavior. That is exceedingly difficult to do when there is no data upon which to base those decisions. For example, I keep searching for data about church reopenings. They’ve been going on for months now. We definitely should have some data that indicate whether reopenings are linked to outbreaks and, if so, what safety precautions they were taking. Did those precautions make any difference? Is there a way to reopen safely? Masks? Physical distance? We’ve heard lots of anecdotal evidence, but that isn’t really data.
A study was released today that indicates of people who test positive for COVID19, twice as many say they went to restaurants during the two weeks prior to their diagnosis as do symptomatic people who test negative. Lots of questions about the study though. For one thing, using self-reported data is problematic. We don’t know how accurate their reports are. And apparently a number of people in the group refused to cooperate. Also, it is impossible to control for other factors, such as What safety precautions did the restaurants practice? Were patrons dining inside or outside? In what other risky behaviors did they participate? Or maybe I’m just looking for an excuse to go out to eat.
The graphic above is based on New York Times survey data about the number of cases on college campuses. It is a little misleading though. Northwestern has the lowest number of any schools that I have a relationship with, classes hadn’t even started yet when the survey was taken. There’s some other data I’d like to see though. What happened when students at campuses such as UNC and NCSU were sent home? Can we find outbreaks back in their home counties?
People argue over the little bit of data we have and what it means. Is wearing masks the best way to protect others and yourself, or are masks a huge hoax? I don’t think looking at CDC recommendations will help, since they change daily. Perhaps future generations will have a chance to understand what all this means.