Prior to a few months ago, Professor John Jonides (CCN) would never have guessed that he would one day conduct interdisciplinary research into the causes of inappropriate social behavior during a worldwide pandemic. Of course, few people could have predicted the specific social changes COVID-19 has wrought. But Jonides, whose career has largely focused on what he describes as “basic cognitive neuroscience,” would also never have guessed that he would be performing such research because it is simply outside of what he once considered his purview. Indeed, Jonides believes his new research offers valuable lessons on at least two levels. The first is about the potential benefits of targeted, basic scientific education during the COVID-19 pandemic. The second is about how crucial it is for researchers to pursue their sincere interests and intuitions, even when doing so may lead them outside of their previous scientific comfort zones.

Jonides explains that his metamorphosis from “basic” cognitive neuroscientist to quasi-social-psychologist began with a news report. “Back in March, like all of us, I was watching CNN damn near all of the time because the COVID thing was just starting,” he recalls. “On March 18—and I have a vivid memory of this—they showed an aerial view of a beach on the western side of Florida. Kids were on spring break, and they were just gathered on that beach cheek-by-jowl—no masks, no social distancing, nothing. They were just swimming and partying and drinking, and I was thinking, what is going on here? Don’t they know that there is an epidemic and they could get sick by doing this?”

Jonides’s scientific curiosity was immediately piqued. “That observation inspired me to get together with professors Priti Shah and Rick Lewis and with [current and former] graduate students Aysecan Bodoroglu, Madison Quirk, Tyler Adkins, Madison Fansher, and Pia Lalwani,” he says. “We started talking about how people could be engaged in behavior that was so obviously inappropriate given the COVID situation. What kinds of cognitive problems may they have that are causing them to behave that way? Our first intuition was that they just didn’t understand that this was a rising epidemic and that they could be caught in the middle of it.”

To test that intuition, the collaborators presented nearly 900 randomly selected MTurk users with actual information about rising COVID case numbers. Some users received the data in graphs, others in tables, and still others as text in the form of paragraphs. Participants were then asked to project what the case totals would be three days, six days, and nine days from the time of the experiment. What the team found that was that regardless of the format in which the data were presented, an overwhelming majority of subjects—nearly 82%—underestimated the number of future infections, and sometimes wildly so.

The finding was particularly significant, Jonides explains, because at the time, cases were rising at a fairly predictable, nearly exponential rate. In other words, a preliminary finding seemed to indicate that an overwhelming majority of subjects appeared to either not understand or not recognize basic exponential growth as it applied to the pandemic.

As disconcerting as that finding may sound, the second phase of the experiment did offer some encouraging news: namely, that people can improve their predictions with practice—at least in some circumstances. When the researchers asked subjects to make similar predictions a couple of weeks later, their new predictions were both moderately better than their previous ones and better than those of new subjects who had not participated in the previous phase of the experiment. However, that improvement only occurred for subjects who viewed the information in graphical form; those who received the data in tables or paragraphs showed no improvement.

Intrigued, Jonides and his collaborators decided to see if they could help people further improve their predictions with a bit of education. “We launched into a second project where we actively trained people by giving them hypothetical COVID data from hypothetical countries and asking them to predict the future case numbers,” he describes. “We then gave them feedback and said, essentially, ‘No, it’s not like that. It’s actually like this.’ We gave them a few cases. Then we looked to see whether they got better on our post-test, and sure enough, they do get better. It was a modest improvement, but it showed that you actually can help people get better at making these estimations.”

With that information in mind, Jonides says, “We’ve now launched into something far more aggressive. One of our graduate students, Madison Fansher, created an 8-minute video about what it means to flatten the curve. It goes through the entire logic of the number of daily cases and the number of cumulative cases, and it talks about what needs to happen to the daily case curve to flatten the cumulative case curve. We are going to launch an experiment in which we pre-test people on making judgments about flattening the curve, and then we will show them the video. We will then post-test them on whether they’ve learned enough from the video to actually do better.”

The fact that people can improve their predictions is promising, but changing inappropriate behaviors may ultimately require understanding and addressing other psychological factors as well. For example, Jonides speculates that misguided feelings of invincibility among young people may also play a role in their inappropriate behaviors, and those effects may be exacerbated by widespread media messaging about the supposedly lower COVID risk for their demographics. Although those speculations are currently just that—speculations—the researchers are currently looking closely at the data to tease out what other factors may be involved. Ultimately, Jonides hopes these data will help determine which groups would most benefit from interventional education—and which groups most need it in the first place. As Jonides explains, “To change behaviors for the better, you have to know who to target. Just as the internet uses targeted advertising, this may be a way to target training toward the appropriate groups to ultimately help them change their behaviors.”

Again, Jonides argues that his COVID research offers at least two important lessons. First, the studies suggest that a targeted education campaign may well be effective in changing inappropriate social behaviors and ultimately helping to control the spread of the disease. Second, the simple fact that Jonides is the one who initiated these studies illustrates how crucial it is for researchers to pursue their interests and to follow the evidence, wherever that may take them.

“Also,” he adds, “I do have one more bit of advice for students, and it’s about the disease, which is after all, what we really care about right now. Look—just wear your mask, stay separated, wash your hands, and don’t touch your face. If we all did all of those things, we could have this disease under control in no time. Wearing your mask has nothing to do with any abstract political ‘freedom.’ It’s all about social responsibility: you don’t want to give it to others, and you don’t want them to give it to you. So behave responsibly. Act like an adult.”