Emily Oster lives for misinformation. Or, rather, she lives to dispel misinformation. A professor of economics at Brown University and author of Expecting Better: Why the Conventional Pregnancy Wisdom Is Wrong and What You Really Need to Know, Oster has nearly single-handedly brought data analysis to parenting. So it would make sense that, early in the pandemic, Oster co-founded a website called COVID Explained that puts everything you need to know about the coronavirus into context. The site is run by a team of immunologists, economists, biochemists, physicians, and students who work to contextualize information, dispel rumor, and read studies out loud, for all to see how to do just that. Here, Oster talks about how it all came together, what’s next, and what her journey into the heart of COVID misinformation can tell parents about their own research habits. You were seeing data being poorly collected, poorly disseminated, and poorly explained when it was well-collected. How did this lead to the launch of COVID Explained? We launched COVID Explained not because of a concern about data — although, in the end, that’s a lot of what we’ve been doing — but with a general sense that people didn’t understand some very basic facts about the virus. People didn’t understand how you get COVID-19. They either had the impression that this was an imaginary hoax, or they’re at the other extreme, walking around thinking that if someone goes into the grocery store and they have COVID, and then I come four hours later, and I touch the same salad box, I’m going to drop dead right away. With that kind of understanding, it was sort of like, what’s the point in washing my hands if at any moment I could touch a salad box and die? So part of what we were trying to explain was how viruses work. Here’s what your body does to fight them. Here’s why washing your hands matters. Here’s why masks matter. Then there was this idea that a lot of the stuff that’s reported about treatment or about kids or whatever is not given a lot of context. Because we have a bunch of content and facts about the virus on this site, we were able to think about some of these new things in the context of some broader facts. When did you switch to becoming a data collector? That was driven by frustration. It was clear that many of the questions that people have about COVID relate to certain issues they face in their families, particularly as things start to reopen. There were a lot of questions around what should I do with my kids? What should I do with my parents? And I kept thinking, there are places where this data exists. You want to understand what the COVID-19 rates look like in childcare settings that have been open? I kept thinking, pretty soon someone’s going to report that data. That will be so great. We can have a nice page about it at our site. But that data never appeared. Finally, I got tired. We were like, let’s try to get some of this data — with the understanding that our data collection system that relies on Google Forms is not a great system. But there was also the feeling like it would be good to have something. One of the data initiatives we have at the moment is around tracking places over time. We have some childcare centers and camps that have signed up to be in a tracking sample. We ask them every week, do you have any COVID cases? Again, it’s crowdsourcing. But it does give us some sense of what’s happening over time. Maybe we can show that it’s possible to get people to participate, and that might prompt some more official efforts. The COVID Explained site does a really good job of explaining clearly and upfront why this is not official data and what crowdsourced data means. Do you worry that people either don’t read those caveats or, for the media, don’t pass along those caveats and just take your data as an answer? It hasn’t happened as much as I feared. At the beginning, people were like, ‘I can’t believe you’re putting this out there. People are not going to understand this.’ But actually a lot of the media coverage of this has (rightfully) said this is an unscientific survey. We try to caveat it as much as we can, but somebody last week told me the Orange County School District showed these data as a defense of why they should reopen with no masks. That kind of thing obviously makes me sad. It seems like COVID Explained is a breeding ground for testing communication methods. What do you think has been one of the more successful ones? We’ve been experimenting with form a lot and trying to figure out what we can do to communicate that is different from tradition, different from what you would do at a media outlet. Some of it is trying to hit a line in the writing between science, science, science and don’t worry, it’s all fun. So something I work a lot on is how you communicate scientific concepts in a way that is accessible. For example, we put up a conversation that a few of our experts had on Slack. The reason we did that is it’s hard to hear the expert conversations over the stuff that gets out into the world. With a lot of the new information that comes up, it’s really hard to know where to put it in the conversation. Everything is coming as preprints. There are few people in the samples. It’s not consistent with what came out the day before. So in the end, the people who work on the site end up having long conversations in our Slack channel. And I thought that it might be interesting to visualize some of those conversations for people to understand that if you were somebody who really knew a lot about this, you would still be unsure and still be trying to understand what every new piece of evidence says. Whereas we were all hoping it was going to be like a Steven Soderbergh film where you have all these competent people finding answers quickly and efficiently. Exactly. What do you think about changes in how hospitals report COVID-19 data? Two weeks ago, the Trump administration ordered hospitals to stop submitting COVID-19 data to the Centers for Disease Control and Prevention and instead submit it to the Department of Health and Human Services. Already, some of the data has disappeared. Is this a symptom of the problem with data, or is it a secondary problem — something that’s beyond trying to crunch the numbers on a pandemic? Nothing about our data collection in this pandemic has been good at all. We aren’t getting information that we need. We’re not putting together the data we have on testing in a way that’s helpful. We’re not reporting on it in a way that’s helpful. The media is lobbing numbers without providing a lot of context. With the particular issue of how this evidence is reported, one would not generally suggest switching to another reporting system in the middle of a crisis. A lot of people seem to think this is some effort by the White House to hide data. In normal times, I would have said that’s a paranoid fantasy. But I think it’s totally reasonable in this case, given everything else going on. I don’t know how much of it is generic incompetence, and how much of it is active badness. But since our data was already very bad, I’m not sure if this is going to necessarily make it worse. I’m not sure how much worse it could really get. Parents are trying to answer lots of questions now. Do you have advice for them on reading articles on cherry-picked studies, following Twitter threads, doing all this to figure out whether they should send their kid back to school. Or answering other questions, like should I go to Disney World? The main thing is to try to outline upfront what information you are looking for. Then look for it in a more targeted way. I think a lot of what happens is that people are like, okay, let me just google ‘can you get COVID at camp?’ If you Google that, you’re going to get stories saying there’s an outbreak at this campus and this campus — because there are some outbreaks at camp. But if you step back and say, the question I have is ‘what is the risk of my kids getting COVID at camp’ and you recognize that’s going to require seeing how many kids have COVID at camp or how many camps have outbreaks and how many camps are open, it would give you much clearer context for the articles that you are going to see. Because sometimes what happens is people don’t exactly form the question that they want before they start googling, and then it’s easy to get into something really scary or something that’s inappropriately reassuring. Like, 70 people went to Disney and they didn’t all get coronavirus, but that’s probably not the question that you want to know the answer to. Do you have any rules for red flags for parents who see something that seems too good to be true? If you see something that seems very surprising in either direction, you should be cautious. When you see something, that is a moment to check in. Maybe it is something totally different that has happened that we didn’t expect. But take a step back and try to make sure that you really understand what is going on here. Do you have anything reassuring to tell parents right now? Not really. The bulk of our evidence says that kids are not heavily affected and don’t get that sick. So that’s that part remains reassuring. But I feel personally very demoralized about the prospects for schools. So I have nothing reassuring to say about that.
