From Pixels to Preferences: Visual Salience Shapes Economic Choices Beyond Gaze Alone
By integrating eye-tracking with computer vision and a novel image-computable model, this study demonstrates that visual salience significantly shapes economic choices through a gaze-independent pathway, accounting for over 80% of its effect on risky decision-making beyond mere gaze bias.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
Every day, we make countless choices, from picking a snack at the grocery store to deciding whether to invest in a stock. We like to believe these decisions are driven by logic and the actual value of the options before us. However, our brains are not perfectly rational machines; they are easily swayed by how things look. A key concept in understanding this is visual salience, which simply means how much an object stands out from its surroundings. A bright red sign in a dull gray street, or a large font in a block of small text, grabs our attention automatically. For decades, scientists believed that this visual pull worked in a straightforward way: a bright or large item catches our eyes, we look at it longer, and because we are looking at it, we end up valuing it more and choosing it. This idea, often called gaze bias, suggests that our eyes lead our minds, and that if we could control where we look, we could control what we choose.
A team of researchers from the University of Oxford, Zhejiang University, and the University of Pennsylvania set out to test whether this simple story is the whole truth. They wanted to know if visual salience changes our choices only because it moves our eyes, or if it does something deeper and more direct to how our brains process value. To find out, they designed a study where people made decisions about money while their eye movements were carefully tracked. They presented participants with a series of simple gambling choices: a fifty-fifty chance to win a certain amount of money or lose a different amount. The researchers manipulated the visual appearance of these numbers. In some trials, the potential gain was made brighter than the potential loss, and in others, the loss was made brighter. This brightness difference made one number visually pop out, or become salient, without changing the actual math of the gamble.
The results confirmed that people were indeed more likely to accept a gamble when the potential gain was the bright, salient number, and less likely to accept it when the potential loss was bright. This happened even though the actual amounts of money remained exactly the same. The researchers also confirmed that the bright numbers did catch the participants' eyes; people looked at the salient number more often and for longer periods. However, when the team used sophisticated statistical methods to separate the effect of looking from the effect of the visual brightness itself, a surprising picture emerged. They found that the act of looking at the bright number explained only a tiny fraction of the change in decision-making. The vast majority of the influence came from a different source. The visual brightness was altering how the brain internally represented the value of the money itself, making the gain feel larger or the loss feel smaller, regardless of where the eyes were actually pointing.
To understand how this invisible shift in value representation works, the researchers built a computer model that mimics the human decision process. This model took raw images of the gambling options, just like a camera sees them, and processed them through layers of artificial intelligence designed to recognize visual patterns. The model then made a choice based on what it "saw." By peering inside the model's internal calculations, the researchers could see that the visual brightness changed the way the numbers were encoded before any decision was made. It was as if the brightness itself changed the weight of the numbers in the brain's internal ledger, independent of where the eyes were looking. This suggests that visual salience operates through two distinct pathways: one where it moves our eyes, which then slightly amplifies the value of what we see, and a much stronger, separate pathway where the visual properties of the stimulus directly reshape our internal sense of value.
This discovery challenges the long-held view that our eyes are the primary drivers of our preferences. It shows that the way information is presented visually can bypass our conscious attention and directly alter our economic choices. The study involved over one hundred participants making nearly one hundred decisions each, providing a robust dataset that supports these conclusions. The researchers found that the gaze-independent effect was so powerful that it accounted for nearly ninety-five percent of the total influence of visual salience on choice. In other words, the mere fact that a number was brighter changed how people felt about the money, even if they didn't spend much time looking at it. This implies that the design of visual information, from the font size on a menu to the layout of a financial dashboard, has a profound and often hidden power to shape our behavior.
The implications of this work extend beyond the laboratory. It suggests that we cannot simply rely on looking away from a tempting option to avoid being influenced by it. The visual environment itself can rewrite our internal values before we even realize we are making a choice. By revealing that visual salience works largely through a mechanism that is separate from eye movement, the study offers a new understanding of how perception and decision-making are linked. It highlights that the brain does not just passively record what the eyes see; it actively constructs value based on visual cues, sometimes in ways that are completely independent of where we are looking. This dual-pathway view provides a more complete picture of human decision-making, showing that the path from a pixel on a screen to a preference in our mind is far more complex and direct than previously thought.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.