“An apple a day keeps the doctor away.” While the principle behind the proverb—nutritional healing—was championed by the Greek physician Hippocrates over two millennia ago, its wording has also stood the test of time. It is a threat appeal. In communication research, threat appeals refer to a specific message format that identifies a potential danger intended to inspire fear, such as one that would require the intervention of a doctor, and recommends a solution that is supposed to quell the fear: eating healthy. A notable feature of the proverb is its thinly veiled threat. Why not “An apple a day keeps the illness away” or, more bluntly, “An apple a day keeps the grim reaper away?”

The Inverted-U

The question for researchers has been: Are more frightening ones more effective? An old idea called the inverted-U hypothesis says yes, but only up to a certain point. Intense fear distracts people from the cause of the fear, such as a disease, an addiction, or a bad habit, and turns them toward managing fear itself. This can be why someone buries their head in the sand. All that matters is to mitigate the internal turmoil; whether or not one will be harmed by the external threat becomes an afterthought. Therefore, the effectiveness of threat appeals is said to first increase and then decrease as they become more frightening, following the shape of an inverted U.

The inverted-U hypothesis was first proposed over half a century ago and it has been subjected to numerous experimental tests ever since. But, the downward slope of the inverted U has rarely been seen in empirical data. This is often explained by referencing ethics or pragmatics: It would be unethical to seriously frighten people in research settings (so researchers don’t) or it just isn’t possible to create messages that really scare people (so researchers can’t see the predicted downturn).

To illustrate these arguments, look at a study conducted by Hye Jin Yoon and Spencer Tinkham. These advertising researchers conducted an experiment with three versions of an ad for a sunscreen brand intended to induce low, medium, and high levels of fear of sun hazards such as sunburn and skin cancer. As is common, they found the high-fear version to be the most persuasive. But the level of fear reported by participants did not exceed the midpoint of their fear scale in any of the three conditions. These results fit with the ethical argument and with the pragmatic argument. Perhaps the authors intentionally designed their messages to be relatively low fear because they were concerned about the effects of frightening participants. Alternatively, they may have made them as scary as possible, but couldn’t find a way to induce higher levels of fear. Either way, Yoon and Tinkham’s data could not have shown the downward slope even if it truly existed because no one in their study was really, truly afraid.

Rethinking How We Evaluate Fear Levels

We thought that there might be another way to think about fear levels. Instead of the scale midpoint (such as “4” on a scale from 1 to 7), we proposed a different reference point: the level of fear people experience when facing real-world adverse events like war and natural disasters. We tested this idea by assembling studies from the scientific literature into two groups, one of which contained investigations that measured fear induced by adverse events ranging from criminal assaults, terrorist attacks, epidemics, to earthquakes and hurricanes. The other contained studies that measured fear induced by threat appeal messages about diseases, substance abuse, environmental issues, traffic safety, and crime.

To give you an example, a message developed by Davis and Jansen warned about the dangers of chlamydia and recommended condom use and annual STD testing. Because fear can be assessed as the intensity of arousal without referencing specific circumstances (for example, on a scale from 1 [not at all fearful] to 7 [very fearful]), we were able to compare the levels of fear in these two kinds of study. It turns out that the intensity of fear responses was similar in the two types of studies. And this held true regardless of the nature of the adverse events or the topic of the threat appeals. We concluded that threat appeal messages used in persuasion experiments produced similar levels of fear as war and natural disasters like earthquakes and hurricanes.

Now What Do We Think of the Inverted-U?

Because of the severity of the consequences, such as loss of lives and property, the level of fear caused by these extreme events can be considered very high and serves as a ruler for evaluating threat appeal messages. Since these messages induced a similar level of fear, researchers should have demonstrated the downward slope of the inverted-U with them, if it really existed. But they did not: Extreme fear-inducing messages did not produce lower levels of persuasion. Thus, prior research on threat appeals probably did provide a strong test of the inverted-U hypothesis because, by the adverse events standard, research participants in persuasion studies really were quite frightened.

Therefore, given what we know now, there is no reason to expect a downturn in effectiveness at high levels of fear. Rather, it seems safe to say that the effectiveness of threat appeal messages will likely increase as the fear induced by them intensifies.

Everything in moderation may be a useful rule of thumb on many occasions, but not for threat appeals. Next time you want someone to eat healthy, say “An apple a day keeps the grim reaper away.”

For Further Reading

Davis, B., & Jansen, C. (2016). This may come as a surprise: How prior knowledge of information in a fear appeal is associated with message outcomes. Communication42(3), 398-421. https://doi.org/10.1080/02500167.2016.1209536

Dillard, J. P., & Li, S. S. (2020). How scary are threat appeals? Evaluating the intensity of fear in experimental research. Human Communication Research46(1), 1-24. https://doi.org/10.1093/hcr/hqz008

Yoon, H. J., & Tinkham, S. F. (2013). Humorous threat persuasion in advertising: The effects of humor, threat intensity, and issue involvement. Journal of Advertising, 42(1), 30–41. https://doi.org/10.1080/00913367.2012.749082

Shu Scott Li is a Ph.D. candidate in the Department of Communication Arts and Sciences at Penn State University. His research focuses on interpersonal influence, health persuasion campaign, and nutrition intervention for healthy eating.

James Price Dillard is Distinguished Professor of Communication Arts and Sciences at Penn State University. His research revolves around the questions of how messages arouse emotions and the conditions under which those emotions produce persuasion.