Errata Carmona
To Sway Someone’s Decision, Use Numbers
A grade of 97 percent looks better than an A.
- By
- April 29, 2025
- CBR - Behavioral Science
Say you’re trying to decide between two cars, one that’s more environmentally friendly and the other with a better safety score. Which one do you pick?
People tend to choose on the basis of numbers rather than words or visuals, according to Toyota Research Institute’s Linda W. Chang, Chicago Booth’s Erika Kirgios, MIT’s Sendhil Mullainathan, and University of Pennsylvania’s Katherine L. Milkman. So if the cars’ safety ratings are described as a 7.5/10 versus a 9.5/10, while their eco-friendliness is depicted with a graph—you’re more likely to choose the safer car. But flip the script and see eco-friendliness quantified instead and you’re likely to go for the greener option.
The same holds true for all sorts of decisions. For example, an employer struggling to pick between two strong candidates—one with a higher GPA and one with more relevant work experience—might be more likely to choose the applicant with the higher GPA because that’s more easily quantifiable than work experience. This “quantification fixation” doesn’t always lead to the best decisions—a higher GPA doesn’t mean a candidate is right for the job—but it does indicate that numbers may sway decision-makers in trade-off scenarios.
The researchers conducted 21 experiments with more than 23,000 participants recruited from online sites, social media, college campuses, and pop-up labs. These experiments demonstrated decision-making patterns consistent with quantification fixation in a variety of situations, including choosing between job and summer internship candidates and selecting charity donation recipients. When facing trade-offs, people tend to gravitate toward whichever option dominates on features presented numerically, the researchers find. “When we count, we change what counts,” they conclude.
In the experiment involving a choice between two hypothetical summer internship candidates, the researchers conveyed information about the candidates’ grades randomly in two ways. They told some participants that the first candidate had a calculus grade of A and a management grade of 83–87 percent, and that the other earned a B in calculus and a 93–97 percent in management. They told a separate group of participants that the first candidate had a calculus grade of 93–97 percent and a B in management, and that the other had an 83–87 percent in calculus and an A in management.
The same information gained prominence when it was expressed as a number. For example, 84 percent of participants chose the candidate whose management grade was represented as 93–97 percent, while only 69 percent chose the same candidate when the management grade was instead described as an A.
“These results suggest that even in a setting where the numeric and non-numeric information provided about a choice are similarly familiar, quantification fixation persists,” the researchers write.
The power of numbers
In the charities experiment, the researchers were able to test whether quantification fixation arises when in-person participants make real, consequential decisions. Participants received detailed information from an independent auditor about two charities, the Natural Resources Defense Fund and the Nature Conservancy, and decided which they would donate $1 to. The auditor scored both organizations along two key areas: accountability and finance and culture and community.
As in the internship experiment, the scores were represented as either numbers or graphics. There were two experimental conditions, and in both cases, the NRDF’s accountability and finance scores were higher than the Nature Conservancy’s, but its culture and community scores were lower. The only variation across conditions was which score was presented numerically and which was presented graphically.
More participants donated to the NRDF when its accountability and finance score was presented numerically, 57 percent compared with the 41 percent who gave when the same equivalent score was shown on a graph.
“In other words,” says Kirgios, “whichever charity dominates on the dimension presented numerically is preferred—that is, we get a ‘preference reversal.’ Nothing else changes across conditions: These are the same real charities, and we’re presenting exactly the same information about them. But people treat them differently based on how that information is shared.”
People’s tendency to overweight numeric information is driven by the ease of comparing numbers, the researchers explain. When facing trade-offs, it’s natural to consider which option is better on each dimension, and by how much. Numbers make such comparisons feel easy and intuitive, even relative to verbal descriptions, visuals, and graphs. As a result, people—especially those who are more comfortable with numbers—feel more confident relying on numeric information to make decisions.
The researchers note that there are many important factors in decision-making that are not easily quantified. For example, when deciding whether to receive treatment for terminal cancer, the trade-off may be life expectancy (easily quantifiable) versus quality of life (harder to quantify).
Understanding that people gravitate toward numbers when facing difficult trade-offs will help those, such as doctors, who are in the position to present the options. If a presenter can’t quantify one choice, they should make an effort to stress its magnitude so decision-makers are not “pulled away from valuable qualitative information toward potentially less diagnostic numeric information,” the researchers write.
Linda W. Chang, Erika Kirgios, Sendhil Mullainathan, and Katherine L. Milkman, “Does Counting Change What Counts? Quantification Fixation Biases Decision-Making,” Proceedings of the National Academy of Sciences, September 2024.
Your Privacy
We want to demonstrate our commitment to your privacy. Please review Chicago Booth's privacy notice, which provides information explaining how and why we collect particular information when you visit our website.