We need to talk about John Snow. (No, not the Game of Thrones character, that’s Jon Snow.)
John Snow was a physician in 19th century England. In his day, a terrible disease was ravaging Great Britain. Off and on, cholera would break out, causing hundreds to become acutely ill with diarrhoea and vomiting. Many died within days of being infected.
The dominant theory of what caused it? “Miasma” – stinky air. It made sense, since the affected places tended to smell terribly. But Snow had a different idea: he believed that cholera spread through water rather than air, and he was looking for evidence to prove that.
1854 brought a new outbreak in the London neighbourhood of Soho. Snow started walking down Soho’s streets and canvassing the locals. He noted where the deceased had lived, and later mapped his data on a map, using black bars to represent cholera deaths. Most of the bars were close to a specific point: a water pump in Broad Street. Snow’s findings matched his theory, prompting the local authority to remove the handle from the pump, so people couldn’t drink from it anymore. Shortly afterwards, the outbreak petered out.
With his map, John Snow had become one of the pioneers of the infographic, and showed how data can save lives. Numbers played a central role in his study. By counting the deaths and visualising the numbers this way, he made a compelling case for acting swiftly.
It is no wonder that today, in the battle against Covid-19, we are once again turning to numbers. For months we have heard daily bulletins about new cases, official guidelines citing the need for "flattening the curve", and warnings about exponential growth. And our timelines have been teeming with posts by your cousin’s neighbour’s brother showcasing his own calculations of where the pandemic is headed. I’ve been writing about numeracy for years, and never have I seen interest in numbers grow so rampant so suddenly. The World Cup, sure. Elections, definitely. But nothing releases the inner nerd like a pandemic. For all of us, numbers offer some certainty in uncertain times.
As powerful as they may be, coronavirus figures also have their limitations. From the very start, mistakes were made by both laymen and experts. Let me mention three emblematic examples.
First: misleadingly precise numbers. It was immediately clear that the actual coronavirus cases were much higher than those reported in official figures, and circulated, often uncritically, on social media. Since in most countries not everyone was getting tested, the figures could at best gauge the lower bound of the actual numbers.
Second: skewed samples. For instance, a study by Stanford researchers to estimate the prevalence of the virus in Santa Clara County, California, concluded that Covid-19 was much more widespread than initially believed. Researchers had used Facebook ads to recruit test subjects, which is a convenient method, but also one that distorts the numbers, as people who feel sick are likely to self-select into such a study in order to get tested.
Third: mistaking correlation for causation. There were endless speculations about which countries did “best” in tackling the pandemic. It is tempting to gaze at the figures and come up with some facile explanation for the differences. Some have claimed masks are the solution, others that we should all have robust contact-tracing like South Korea or a laissez-faire approach like Sweden. But there is no silver-bullet solution. Countries have different testing strategies, different demographics, different healthcare systems. And some countries simply had more luck, as the pandemic knocked on their door later and gave them more time to prepare. How countries fared probably depends on a complex interaction of many factors, which will take years to disentangle.
So how can we make sense of this crisis? Obviously, numeracy helps. But there is another important ingredient: embracing uncertainty. There were so many questions that were unanswered at the beginning of the crisis. Do facemasks help? Does the virus survive on surfaces? Are you immune once you have antibodies? Around the world, scientists are piecing the evidence together to get answers. The way they go about it is by recognising uncertainty.
John Snow – now often considered the father of epidemiology – teaches us how to doubt productively. He was bold enough to question the miasma theory. But he was also wise enough not to be satisfied when he found the Broad Street pump. It could have been the case, he thought, that the air around the pump was particularly smelly, and that the miasma was indeed the culprit. Correlation, after all, is not causation. So he searched for more evidence.
He found a brewery that was located close to the pump, but almost none of the workers had contracted cholera. And then he heard of a woman who lived a few miles away from the pump but had still contracted cholera. Both instances seemed to jar with Snow’s theory. On closer inspection, though, they turned out to be just the clinchers he needed. The brewers, it turned out, had been saved by the beer: when they were thirsty, they drank the brewery’s ale (and, otherwise, water from the company’s own well). As for the woman, she had had a daily delivery from the Broad Street pump, because she liked the taste of the water.
Even later, when people claimed that the outbreak stopped because the offending pump’s handle had been removed, Snow kept doubting. He actually pointed to the fact that many people had fled the neighbourhood, probably causing the outbreak to stop regardless of the intervention on the pump.
Snow shows that doubt makes you stronger. Only by continuously questioning the evidence, you get closer to the truth. Thus, during the pandemic, I’ve been looking up to people who dare to doubt. I listen to the journalist that recognises the uncertainty in the world, the scientist who is not ashamed to change his mind, the friend who collects evidence first and only then forms her opinion. Recognising uncertainty is the best way to interpret numbers, and information in general. I have many doubts, but I’m pretty certain of that.
Sanne Blauw is the numeracy correspondent for De Correspondent and author of The Number Bias: How Numbers Lead and Mislead Us
This article was originally published by WIRED UK
