I’m no data scientist but I’ve been tracking black folks and the coronavirus since March (BlackCovid19.com) and it’s been obvious from the beginning that COVID must have a devastating impact on mankind’s ability to count numbers correctly.
All of a sudden, nobody can produce good numbers. The formulas change almost daily. You ask for numbers and the reply is a bunch of words.
After a while, it became obvious to me this new example of bad math is by design. I’ll probably never know why organizations who have been so good at capturing and disseminating numbers forgot how to do it all of a sudden.
Today article in the Philadelphia Inquirer revels just how bad the coronavirus number crunchers are (Health officials failed to safeguard trust in Pa.’s coronavirus data, sowing confusion.)
It makes no sense for me to try to explain it, so here are a few quotes from the article that should encourage you to read more if you are as baffled at coronavirus arithmetic ignorance as I am.
Data is a tool for sound policy. Data can cut through noise to help spot trends, like new clusters of the coronavirus. It can prove which communities face the most risk and where supplies are most needed.
They still haven’t figured out how to count black and brown people who tested positive for COVID-19. Could it be because a true count ‘can prove which communities face the most risk and where supplies are most needed?’
…over the past three months, the state Health Department has repeatedly failed to safeguard the public’s trust in its data. While there is no evidence of intentional manipulation, the state’s blunders have created openings for confusion.
The state has not always clearly explained to the public what numbers mean, what they don’t, and when and why they might change.
When mistakes have been made, the state has quietly edited information without clearly documenting and communicating the changes.
The state has been opaque about its methods and sources used to compile the data, obscuring the public’s ability to scrutinize the numbers.
Taken together, these shortcomings make it difficult for researchers, policymakers, and the public to get an accurate sense of what’s happening.
That’s just from the first couple paragraphs of the article. The details read like a horror story.
I had one person question the numbers used in my blog posts on Chester COVID-19 cases which uses number brought to us by the Chester County Department of Health. It was refreshing to know someone is paying attention to the numbers and wonder what they all mean. The exchange went like this…
COVID-19 deaths will never be correctly reported because folks can’t even decide what a COVID-19 death is. Infected folks who never saw a doctor or took a test and survived by roughing it out at home will never be counted. A person who moves out of their home to reside in a nursing home becomes counted as a resident of the nursing home town. But, does an inmate in the county jail who catches COVID-19 get counted as a Concord, PA resident? When they get released, does their COVID-19 positive case count follow them to their hometown?
Coronavirus arithmetic is flawed. Some of it is on purpose. Some of it is because no one is setting rules. Don’t take any of these numbers as gospel.