Every now and then, I remember the Slideshow. This presentation about sexually transmitted infections was infamous among my fellow South Florida seventh graders. If you got your middle school sex ed elsewhere, it might be hard to imagine just how graphic its slides were. But being gross was the point: If kids saw the symptoms of untreated gonorrhea, the reasoning went, then disgust would sway them from reckless decisions.

Down in the basic wiring of our brains, disgust motivates avoidance. You’re less likely to go on a second date with a first date who smells bad. If a pigeon picks at your sandwich, you might opt to go hungry. Public health data backs this up: When cigarette packaging shows graphic pictures of smokers’ diseased organs, attempts to quit smoking double. “A vivid image is much more powerful than just abstract numbers,” says Woo-kyoung Ahn, a professor of psychology at Yale University. “Disgust is a powerful emotion rooted as an evolutionary adaptation that helps us expel and avoid harmful substances.”

So last year, in the depths of the Covid pandemic, Ahn and one of her undergraduate students decided to study whether disgusting images could affect people’s attitudes about infectious disease. They presented graphic Covid-related pictures to unvaccinated people and measured whether seeing those images changed the subjects’ stated willingness to comply with public health advice, such as getting vaccinated, masking, and social distancing. The pictures were rough: swollen Covid toes and dark, necrotic lungs. “Yeah, this is really bad,” says Ahn, looking back at the images.

The disgusting images significantly increased some people’s stated willingness to get vaccinated—more than “non-disgusting” Covid pictures, like physicians wearing masks or patients receiving vaccines—and more than headlines about vaccination incentives, like campaigns that gave free doughnuts or ran lotteries for people who could show a vaccine card. And while they didn’t see much of a difference among people who consider themselves liberal, they did see an effect among self-identified conservatives. The team published their results (titled “It’s Time to Be Disgusting About Covid-19”) in May in the journal PLOS One. Health officials are always looking for ways to influence public behavior for the better—and maybe they don’t employ graphic depictions of reality enough, Ahn posits.

But the task of turning bad images into good behavior is a tricky one. “It's important to do work like this—to attempt to make people more cooperative, more pro-health,” says Paul Rozin, a professor of psychology at the University of Pennsylvania not involved in the work. The study’s result is somewhat expected, given what scientists know about the neurological and psychological effect of disgust. Yet it remains unclear how well that would translate to action. As powerful as disgust is, human behavior may be too stubborn for it to really work, Rozin says: “Most people are immune to almost everything they see.”

Disgust creeps beyond revolting wounds and rotten food. The brain’s insular cortex, which is associated with bodily disgust, also lights up for what’s called moral disgust—stuff like images showing drug addiction. And moral disgust shapes how people interact with each other. We might give a wide berth to someone rumored to have done something distasteful, just as we would to someone with a hacking cough.

Previous studies have found that people who identify as conservatives are more sensitive to disgust. Some researchers have argued that this connection rationalizes social exclusion and ethnocentrism on the basis of “moral disgust” and that, in that sense, a propensity for homophobia, germophobia, and xenophobia actually run together.

Ahn’s own research deals with finding factors that influence reasoning. While you might think that a person who is more sensitive to disgust would be more inclined to follow public health guidance, that hasn’t been the case with vaccinations for Covid. During the Trump administration, many conservative leaders, starting with the president, downplayed its risk, likening it to a simple cold or a flu while criticizing public health measures like mask-wearing and the closures of business and schools. Some of those same parties have tended to greatly overstate the risks of getting vaccinated. While clearly false, vaccine myths persisted, a difference that can now be seen in the divergent vaccination and death rates in red and blue states.

For her study, Ahn wanted to test whether what worked for anti-tobacco campaigns—like those images of bloody oral cancers—could help overcome this partisan split. She and her student, Kellen Mermin-Bunnell, posed their first questionnaires to online study participants in late February 2021. (By this point, the US had begun Covid vaccine rollout primarily to older people, the immune compromised, and essential workers.) Nearly 400 people shared their political affiliations and ranked themselves from strongly liberal to strongly conservative. Some were randomly chosen to see five pairs of disgusting images and headlines related to Covid. The rest saw a much more ordinary set of Covid-related photos.

Then the researchers asked the participants to rate their likelihood of following Covid guidelines from the US Centers for Disease Control and Prevention—like wearing masks, social distancing, and getting vaccinated—yielding an overall “compliance score.” For the group that identified as conservative, this average score was about 72 for those who saw gross images and about 60 for those who saw tame ones—a fairly large gap. For those who identified as liberals, the scores were around 85 and 88, respectively. This wasn’t a big difference. In fact, you might have noticed that the score was actually slightly lower for people who saw the gross images, but Ahn says it’s not enough to be statistically meaningful.

When asked specifically about willingness to get vaccinated, among conservatives the average score was 63 for those who saw disgusting images and about 53 for those who saw ordinary ones—again, a notable gap. Among the liberals, the averages were about 83 and 87, respectively (again, a small split that Ahn says isn’t statistically significant).

The second phase of the study was conducted in early April 2021 with 1,500 people, when vaccines were more available. Because vaccination incentives—perks like money and weed—had also been rolled out by this time, the team added a third variable to their study: whether seeing headlines about these bonuses would motivate willingness to take precautions.

In this phase, scores among the liberals—both overall and for vaccine willingness—remained consistently in the 80s, no matter what. Ahn thinks the stagnant number could mean that liberal compliance was already at a “ceiling,” a high past which it couldn’t improve. Or, perhaps, they were in fact less sensitive to disgust.

But among conservatives, cuing disgust changed intentions more than showing people news stories about incentives or showing benign images. The overall compliance score was about 65 among those who saw disgusting images—8 points higher than those who saw tame photos and 9 points higher than those who saw headlines about incentives. On willingness to vaccinate, conservatives’ average score was about 55 for those who saw gross photos, 39 for those who saw regular photos, and 44 for those who learned about vaccination incentives.

“There’s something about the concreteness,” Ahn says of graphic imagery. She thinks that images could be particularly useful when deployed to nudge people at a precise moment, like putting posters inside public venues: “When there's a sign for ‘Please wear masks,’ there could be a picture” of diseased toes or lungs.

But there’s a big X factor: Nobody knows how long the effects of disgust last. Ahn’s team didn’t test whether the participants in their study actually did get vaccinated later or if their masking or social distancing behavior changed.

Rozin suspects the feelings fade. About 10 years ago, he conducted a similar study on freshmen and sophomores in his Intro to Psych class. He had the freshmen read The Omnivore's Dilemma, a book about the food industry that challenges the business and ethics of eating meat. The sophomores didn’t have to read it. And when asked, the freshmen showed more concern about eating meat and trusting agricultural corporations. “It did have an effect, but it didn't last,” says Rozin. The following year, those same students’ self-reported concerns about the food industry fell to match those of newly-arrived freshmen who hadn't read the book. “This was reading a whole book—a really good book—and having a session with faculty members talking about it,” he says, which should be more persuasive than just seeing a few images.

It’s also hard to know which images might be the most persuasive. For example, violent images have often been used to show the public the human cost of war. “In the Vietnam War, that picture of the person being shot on the street had a powerful effect,” says Rozin, referring to a photo of the execution of Nguyễn Văn Lém. “There were lots of other gory pictures that didn't. But some pictures become iconic. We don't know how that happens. But it does happen.”

In the wake of mass shootings, viral infographics and data have undoubtedly helped rally public opinion for gun control. “Numbers don't lie,” says Eric Patrick, who studies information design at Northwestern University. But, he says, “I think we've peaked with infographics and information design.” Perhaps visually displaying the true toll of gun violence would work, he says, but he’s not entirely convinced that it’d be worth it—he fears it might further desensitize (or conversely, traumatize) the public.

For diseases, graphic images certainly help communicate the symptoms people should look out for. That’s true with the current outbreak of monkeypox, for example. Photos of the sores it causes have exacerbated fear of its spread and might make people extra cautious. But health communicators want to be sure that aggressive messaging doesn’t morph into stigmas. As with HIV, early cases of monkeypox have skewed toward men who have sex with men, but all demographics are vulnerable, as the disease spreads through skin-to-skin contact, handling shared objects, and respiration. Disgust-based messaging campaigns could wrongly exacerbate social divisions. “You don't want to go down a path that’s going to lead to stigmatization,” says Glen Nowak, a professor of advertising and health communication at the University of Georgia.

There are also practical challenges to consider for any campaign. “How people perceive it, and how they respond, is beyond your control,” says Nowak, who worked at the CDC for 14 years, including six as communications director for the national immunization program. “Say you put a picture of people throwing dirt on a grave, like ‘Covid can kill.’ You're probably going to get a lot of people who are going to be upset.” This messaging risks backlash, especially if it comes from taxpayer-funded organizations.

In fact, the data on what might convince people to get vaccinated is all over the map. It varies widely over time and among different groups of people—as do their reasons for not getting vaccinated. (In 2021, the polling company Civis tested whether any of eight types of messages convinced unvaccinated people to want the shot. They found that the most graphic option—deathbed testimony from an unvaccinated person—actually made people less likely to want to get vaccinated. The message that fared best among conservatives was one of personal responsibility. And, overall, the message that worked best in 2021—getting vaccinated to protect kids—hadn’t polled well the previous year.)

The problem, says Nowak, is that not everyone will react the same way. That’s true in Ahn’s study, in which gross images didn’t seem to persuade liberals or all conservatives. And that’s been true for anti-smoking campaigns that rely on graphic images, which have shown that some people just won't be swayed, even by decades of attempts. “They are going to find ways to cope with those images,” says Nowak. “Ignore them. Flip the pack around so they don't see them. They'll say, ‘Yeah, but that's not me.’”