In brief: Unless you understand a study, and believe it to of sufficient quality, do not post the study or news stories about it on your social media. Let’s all do our part to stamp out science gossip, the misinformation and fear mongering about coronavirus (and other matters) on social media.
I must thank my dear friend Danielle for sending me the link to this Coursera course called Science Literacy. It was nice to get a refresher on the basics of understanding what makes a good study, and how to determine whether a news report is an accurate representation of said report (spoiler alert: more often than not, it isn’t). I passed, by the way. Here’s the proof:
It’s free to take, by the way (the certificate will cost extra). And it’s definitely worth the time. There’s a lot in that course that will help you with analyzing the avalanche of information you’re seeing these days.
It was nice to know that I was instinctively doing one thing before this course — questioning studies. In a previous blog post, I talked about an Australian study that found the coronavirus could survive on polymer bills for four weeks, which predictably sent many into a new tizzy about the virus lurking on every surface. At the time, I asked whether the study used unrealistically high levels of virus? I couldn’t answer, because 1) I am not an epidemiologist, so am not qualified to analyze that study; and 2) didn’t have an original copy of the study, and was relying on news reports.
Turns out, others more qualified than me shared the same concern:
The Australian paper was a greatest-hits compilation of research errors, [Emanuel Goldman, a microbiology professor at Rutgers New Jersey Medical School] told me. “It used so many unrealistic conditions to favor the virus’s survival,” he said. “They kept samples in the dark to spare it from light that kills the virus. They used optimal temperature and humidity.” Most important, their samples included a gargantuan amount of virus—“thousands and thousands of virus particles, when the research on influenza indicates that coughs or sneezes emit something in the range of 10 to 100 virus particles in a droplet.”
It’s nice to be vindicated.
But vindication aside, the damage is done. The study with the scary stat has made its way around the world. No amount of correction will change that, nor will it help dispel the rather distorted information that both the study and the news reports have already disseminated around the world. This wouldn’t be so bad if it were an isolated incident. But we know it’s not. There was this story that made me stop following one news organization with the scary headline Coronavirus mutation emerges that may outmaneuver mask-wearing and hand-washing. You guessed it — it was a study lacking peer review, a vital step before publication in a journal.
How about a non-coronavirus example? Recently, we were told that our Netflix streaming habits was killing the planet.
In July 2019, the Shift Project released a report titled “Unsustainable and Growing Impact” of online video. The report claimed that video streaming was responsible for more than 300 million tons of CO2 in 2018, roughly equivalent to the emissions profile of France.
In July 2019, the Shift Project released a report titled “Unsustainable and Growing Impact” of online video. The report claimed that video streaming was responsible for more than 300 million tons of CO2 in 2018, roughly equivalent to the emissions profile of France.
That implies that binge watching Tiger King on Netflix created as much carbon emission as a car that traveled a distance of 2,000 kilometers.
That’s a scary stat. But it also wasn’t correct:
Sometimes, the problem with “think tanks” is that they look for data to support a conclusion rather than drawing a conclusion from the data. In this case, the Shift Project over-estimated the carbon emissions from watching live streaming over the Internet by a factor of 80.
Still, the damage was done. Again. People will remember the original report a lot more than the correction. (See this video from Simon Clark, a PhD in climate science, for a wonderful explanation on the problem with these kinds of stories. His whole channel is really good, so I recommend giving him a follow on YouTube. This is not a paid promotion.)
I don’t think I need to list other examples of scary headlines based on flawed or erroneous interpretations of science. But there’s a bigger issue, as the Atlantic article points out:
The cliché is that people should “follow the science” and do whatever “science says.” But the truth is that science says many things at once. Science says that the coronavirus can last one month on surfaces; it also says it’s vanishingly rare to get the coronavirus from surfaces. Bad studies, good studies, and mediocre studies are all part of the cacophonous hydra of “science” that is constantly “saying” stuff. In the telephone game between scientific research, media reports, and public understanding, dubious studies get simplified, exaggerated, and concretized as gospel.
So bad study leads to bad reporting. But there’s another step, one you can help with: social media posts and analyses that are even more flawed. I’ve lost count as to the number of social media posts that are proclaiming the end of times with every new study. And the analysis was, most often, not accurate. It’s okay — most of us aren’t. But posting studies that we don’t understand, or worse, misinformed stories about studies we don’t understand, does not help. It makes everything worse, in fact. Even worse is when people post one study that claims it refutes established science. One study does not overturn decades of previous studies. That one study, if strong enough, means we need to revisit what we have, but it does not by itself overturn anything. But posting a flawed study and claiming it rewrites what we know? That’s irresponsible.
All of this is what I call science gossip — the repeating of unvetted studies, often with distorted analysis or simply incorrect claims. And we’ve all engaged in it. Scientists writing an exaggerated media release, the journalists lacking qualifications to report on science stories, and all of us who have shared stories or bad studies. Science needs to be able to filter out the bad studies, and look at the strong ones and see if they’re replicable, and what the bigger implications are. Science gossip detracts and distracts from this process. And right now, we need science to be able to do its thing more than ever. (Remember hydroxychloroquine? Donald Trump remembers. As do all the people who wasted time studying this drug because of his “influence” on the matter.) The point: Science gossip spreads misinformation and, in extreme cases, can redirect resources to fruitless endeavours.
Clark brings up another serious problem with studies that provide eye-catching stats: They create a culture of defeatism. If things are this bad, why bother? Let’s just keep doing what we’re doing, we’re all going to die anyway. Of course, there’s a parallel to coronavirus. I recall seeing a story on Facebook a while back that described a study (!) that said immunity appears to be long-lasting. Immediately, the story was flooded with comments about how no, you don’t get immunity from infection. Of course, we don’t know how long immunity lasts, only that there does appear to be some sort of immunity (although it seems to vary from person to person, as various studies seem to indicate). And this is important, because the whole basis of the vaccine is that we can get immunity. So why bother with a vaccine that might cause some side effects if we’re never going to get immunity from it? Why bother with lockdowns? Why can’t we just live our lives, since we’re all going to die?
Science gossip can have very real, damaging consequences.
So how do we deal with science gossip? I propose we all take part in the Post-No-Study Challenge. The concept is simple: Post no study that you haven’t read and analyzed. Have you determined that this study has sufficient controls in place to prevent distorted responses? Was the sample size appropriate? Have you identified study shortcomings? How about potential biases or conflicts of interest with the study authors? If you can’t do that with confidence, don’t post the study. And certainly don’t post a news report of a study you haven’t read.
Why should we do this? To help cut down on the science gossip. We will never be able to eliminate it, because for many people and organizations, there’s money to be made. But we can at least limit the spread of science gossip within our circles.
Let’s wait for science to come to some sort of conclusion based on all the evidence we can gather before posting any studies on the matter, especially if we don’t understand them or haven’t identified flaws in the methodology or assumptions. All links from the Daily Mail are right out. Others are going to spread science gossip. We don’t have to. (And it might even do your mental health some good.)
This does mean I won’t be able to comment on those news stories full of idiots and trolls spewing misinformation on Facebook. But now might be a good time to start unfollowing some news sites, anyway. Lord knows my mental health can use all the help it can get.
Take the Post-No-Study Challenge for 30 days. Let’s do our part to let science do its thing without the interference of science gossip.
Canuckistan Takes Paris! 