A few years ago, Jevin West told fellow University of Washington professor Carl Bergstrom that he was starting a new course on big data. “Oh yeah,” Bergstrom joked, “I’m starting a course called ‘Calling bullshit on big data.’”

The pair worked together to develop a course, Calling Bullshit, broadening the scope to offer tips on how to detect and disarm spurious appeals to data and science in anything from TED talks to medical papers. The syllabus went viral, and dozens of universities around the world now draw on the UW material. Bergstrom and West reoriented their careers around bullshit detection, wrote a forthcoming book, and in December established a new Center for an Informed Public.

A month later, the novel coronavirus arrived. The professors quickly realized it would be their toughest assignment yet in forensic scatology. The pandemic has added Miracle-Gro to what Bergstrom and West’s course calls the “natural ecology of bullshit.” Human nature and society—particularly online—offer psychological and monetary rewards for attracting attention, regardless of whether information is accurate. That the president of the United States has repeatedly spread untruths about the coronavirus and the government’s response aggravates the situation.

As the virus spreads, Bergstrom and West have been deluged by calls for help checking suspect claims and have helped clean up Covid-19 misinformation on Twitter and elsewhere. What they’ve found offers tips on spotting and avoiding the information hazards of pandemic times—and suggests we will be navigating them for a while.


Last week, Bergstrom, an infectious disease specialist, chased down a viral Twitter thread in which neurologist Scott Mintzer at Thomas Jefferson University in Philadelphia described ICUs in Seattle, 2,800 miles away, flooded with dying patients, citing an unidentified intensivist. It included a claim that doctors were withholding life-saving equipment from overweight patients.

Bergstrom reached out to Seattle health care workers to check on the claims and received many messages refuting them. He also contacted Mintzer, who he says initially defended his posts but later deleted them. “The lesson is to be really questioning of unsourced, second-hand reports when they are the most shocking and dramatic,” Bergstrom says. Mintzer told WIRED, “Disputes about accuracy were not the main reason I took it down."

University of Washington professor Kate Starbird used a database of tweets about Covid-19 to create this chart showing how retweets (blue circles), quotes (orange diamonds), or retweets of quotes (green circles), boosted a tweet sharing inaccurate scientific claims about the novel coronavirus.

Courtesy of Kate Starbird

Saturday evening, Bergstrom spent several hours writing a 31-tweet thread debunking a widely shared Medium post by a tech worker who said, based in part on his experience in viral marketing, that officials are overreacting with strict limits on personal movement. Soon after, Medium replaced the post with a message stating that it is “under investigation or was found in violation of the Medium Rules.”

Advertisement

Some conservative figures have rallied around the deleted Medium post, which the author has reposted elsewhere, saying its removal unfairly quashed his freedom to express his opinions. Bergstrom says his debunking of the post and intervention on the Seattle ICU thread have triggered waves of online abuse. He believes his fact-checking attempts have been flagged on trolling forums, such as 4chan. West says preliminary analysis of a database of coronavirus-related tweets that UW has gathered since mid-January suggests right-wing trolls are actively boosting Covid-19 misinformation.

Once information has been amplified online, West says, it can come to be seen as authoritative even by people who should know better. Earlier this month, a friend asked him to help Washington state dentists debating whether to shut their doors to patients. One of the main pieces of evidence shaping the discussion was a Medium post in which a marketing executive without public health expertise warned of the seriousness of the pandemic and offered charts and models he had made predicting its future growth. “These were medical professionals, some of the most at risk-workers, sharing advice from an inexpert source,” West says. The state’s governor shut down non-emergency dental appointments on March 19.

The Covid-19 crisis has also unleashed a plague of beguiling but confusing or even misleading charts and maps. Bergstrom has pushed back on one widely trafficked apples-to-oranges comparison that made the rapidly growing death toll from coronavirus look trifling by placing it next to the larger but steady fatalities from endemic diseases such as malaria.

How Does the Coronavirus Spread? (And Other Covid-19 FAQs)

Plus: What it means to “flatten the curve,” and everything else you need to know about the coronavirus.

Catherine D'Ignazio, an MIT professor and coauthor of the recent book Data Feminism, advises caution when viewing Covid-19 visualizations of any kind, even from authoritative sources like the Centers for Disease Control. The agency offers a map displaying the number of coronavirus cases by state in ranges, using shading. That format, known as a choropleth map, means high-population states like California will appear worse-hit than smaller states even if they have lower rates of infection, D’Ignazio says.

Many charts and maps don't attempt to convey the huge uncertainties in Covid-19–related data, caused by problems ramping up testing, particularly in the US. One Ohio official recently told WIRED her state’s count of cases was likely wrong by a factor of more than 10,000. “Data visualization carries the aura of certainty—clean lines and geometric shapes and reputable sources of data all convey authority,” D’Ignazio says. “But in situations like this, those conventions are doing us a disservice.” She notes that immigrants, women, and low-income people are more likely to be among those missing cases because they are less likely to be willing or able to seek testing.

Despite the bullshit bonanza, West says he has been pleased to see medical experts fighting incorrect information on social networks and tech companies such as Facebook and Google adding banners and filters to fight or block coronavirus misinformation. That suggests these companies could do more to combat other forms of spurious content, such as around elections, West says. “They’ve proven they can do more,” he says.

Bergstrom says the best way to improve your Covid-19 information diet right now is to—as with the SARS-nCoV-2 virus itself—minimize your exposure. Unlike in a crisis of imminent physical danger, like a natural disaster or war zone, staying updated on a five-minute cadence simply isn’t necessary, he says.

“This is a crisis unfolding in slow motion, in a statistical way where we can only see pieces of it,” he says. “I recommend people pick one maybe two times a day to read what’s going on from reputable sources like The New York Times or STAT or WIRED—and if you must go on Twitter, block the hashtags.”

Now send this article to your friends, log off for a few hours, and wash your hands.


WIRED is providing unlimited free access to stories about the coronavirus pandemic. Sign up for our Coronavirus Update to get the latest in your inbox.


More From WIRED on Covid-19

LEAVE A REPLY

Please enter your comment!
Please enter your name here