Throughout the Covid-19 pandemic, health officials updated the public on the outbreak through statistics — case counts, vaccination rates, test distribution. Whether displayed through graphs, charts, or interactive visualizations, these numbers are meant to help the public make decisions in response to health risks.
But do these statistics actually change individuals’ perceptions of risk and behavioral decisions? A new study from Annenberg Public Policy Center (APPC) researchers Dolores Albarracín and Haesung Annie Jung found that they do, but some more than others. Their study, “How people use information about changes in infections and disease prevalence,” published this month in Health Psychology, analyzes data on how different information influences people’s perceptions and decisions during a pandemic.
Albarracín is the director of the Annenberg Public Policy Center’s Science of Science Communication division, the Alexandra Heyman Nash Penn Integrates Knowledge (PIK) University Professor at the University of Pennsylvania with joint appointments in the Annenberg School for Communication (ASC), the Department of Family and Community Health at the Penn School of Nursing, and the Psychology Department, and director of the Social Action Lab at ASC. Jung is a research associate at Albarracín’s Social Action Lab.
Jung and Albarracín began their study in summer 2020, when Covid-19 was the third leading cause of death in the U.S. They wanted to know which statistics are the most effective at encouraging individuals to change their behavior — avoid large gatherings, wear a mask daily, isolate when sick, and vaccinate — to reduce their risk for disease. They conducted experiments to analyze the impact of two of the most frequently shared epidemiological metrics of worldwide disease: First, the number of new infections and second, the total number of infections.
Their findings? Information about new infections consistently has a larger influence on people’s decisions to change their behavior than information about the total number of infections (disease prevalence). The impact of prevalence, however, becomes larger when there is no noticeable change in the number of new infections, such that this number is consistently growing or decreasing.
“These two metrics are related, but distinctly important,” Jung says. “The number of new infections signal immediate changes in disease threat. It tells you whether your likelihood of contracting a disease has increased or decreased compared to yesterday, for example. In contrast, the number of total infections signals how common a disease is in a region. This second piece of information is critical in determining how much risk you have if you walk outside without a mask today.”
In other words, she says, disease prevalence communicates absolute risk, such as telling you “you have a 30% risk of getting a disease.” In contrast, increases in infections communicate comparative risk: “You have a 30% higher chance of getting a disease compared to yesterday.” Providing such a reference point allows people to more easily judge whether their risks are lower or higher, so that they can adjust their behavior accordingly.
According to Albarracín, human beings are psychologically sensitive to change, which makes us respond to variations even if these are trivial. If 20 deaths of flu a year become 30 deaths of flu a year, this change is less consequential for becoming infected today than knowing that there are 500 deaths of flu a year right now. However, psychologically, the increase has a greater impact than the absolute value in most cases. Jung hopes this study will encourage the public health communicators to brainstorm new ways to communicate disease prevalence, and in doing so, prepare us for future disease outbreaks.
To read more about the study at the Annenberg School’s website, click here.