What Goes Up: Research Results Seem Larger When Framed as Increases

Scientific findings are everywhere. We see them in the news. They’re cited on food packages and beverage labels. They are discussed in podcasts and argued over by politicians and pundits. And each finding sits within a specific frame.

If researchers discover an intervention that affects how people spend discretionary income, for example, they could describe it as a method that increases savings or decreases spending. A medication could be said to increase attention spans or, conversely, decrease attention lapses. “When people read or hear about a research finding, do they think it’s more noteworthy — for example, bigger — when the result is framed as an increase or decrease?” asks Zakary Tormala, a professor of marketing at Stanford Graduate School of Business. “If there is a difference in how people react to findings based on whether they’re framed as increases or decreases, that could have profound implications for how people perceive and value science.”

In a series of experiments led by doctoral student Courtney Lee and Christopher Bechler, PhD ’21, now an assistant professor at the University of Notre Dame, Tormala tested how people react to findings from real and fictitious studies depending on how the results are framed. They found that framing a scientific finding as an increase boosts its perceived magnitude and makes people consider it more important and more deserving of funding and publication. “We’re not arguing that increase effects are more important than decrease effects,” cautions Lee, “but people seem to perceive them that way.”

Most of the experiments followed a similar format. Participants read about a finding that was framed as an increase (for example: People are more likely to donate to charity after playing Wordle) or a decrease (people are less likely to donate to charity after playing Wordle). Some of these findings were from real research studies. Others, such as the example above, were invented for the experiment.

After reading about a study’s findings, participants were surveyed on a range of measures, from how big they thought the effect was to how much funding they thought should be directed to the project. “We tested this framing effect across a range of fields: social and behavioral sciences, chemistry, physics, biology,” Tormala says. “Across domains, we saw a reliable trend where findings that were framed as increases were perceived as significantly larger, even when we controlled for the exact effect under consideration.” This pattern held across a range of participants, including those who had a PhD.

Participants also reacted differently to results presented as increases. In one study, participants were told about a nonprofit that had successfully identified interview behaviors that either increase or decrease the chances of getting hired. The participants were then told they might receive a $20 bonus, which could be donated in whole or in part to the nonprofit. Those who were in the “increase” condition gave a larger sum than those in the “decrease” condition.

Importantly, the researchers presented the experimental results as directional — increase or decrease — rather than numerical. It’s possible, Tormala notes, that people are less susceptible to the framing effect when they have specific numbers to reference; for example, a decrease of 16% might be perceived as equivalent in magnitude to an increase of 16%. However, prior research, some of it conducted by Tormala, suggests this may not be the case: Even a seemingly fixed difference between two numbers can seem different depending on their context.

It All Adds Up

At the heart of this result, the researchers suspect, is the difficulty of visualizing an absence. “This intuition is rooted in existing research that has compared how people process and perceive addition and subtraction,” Tormala says. “If I have a stack of three paper cups and I add two, then I can clearly see those two. It’s easy to see the addition. But if I have five and remove two, it’s harder to picture what’s missing.” Tormala and his coauthors hypothesize that findings framed as increases are treated as additive, whereas those framed as decreases are more like subtraction. This produces a difference in how clear increases seem relative to decreases.

Clarity, in turn, has a demonstrated relationship with perceived magnitude: Things that are clearer — meaning they feel more concrete and easier to picture — often appear to be larger. Several of their experiments explored this mechanism. Participants who read findings framed as an increase found them clearer than those who read the same findings framed as a decrease. This difference in clarity drove the difference in perceived magnitude.

Even researchers who study behavior appear to be susceptible to the increase framing effect. When Tormala’s team looked at nearly 74,000 articles from 10 behavioral and general science journals, they found that a sizeable majority used increase framing rather than decrease framing. Moreover, articles that used increase framing were cited more often by other researchers.

The results, Tormala noted, are not intended to “help scientists pull a fast one on readers by making their findings seem bigger.” But anyone interested in influencing policymakers or the public, he says, should consider how they present their findings. If you’re arguing for a new education policy, for instance, you may generate more support if you talk about an increase in graduation rates rather than a decrease in dropout rates.

Readers, too, should be aware of how framing can tilt their perception. “There’s a little bit of a cautionary note in here,” Tormala says. “We may overreact to a finding if it’s framed as an increase or underreact if it’s framed as a decrease. For anybody reading a scientific claim, it may be useful to know that our reactions can be shaped by something as simple as word choice.”