Why Visual Appearance Influences Our Trust Decisions

In an era where misinformation spreads rapidly across digital platforms, new research from MIT has uncovered a concerning psychological phenomenon: people judge the truthfulness of visual data based on appearance rather than content. The study reveals that our assessment of charts and graphs is heavily influenced by visual cues that trigger snap judgments about the creator’s intentions and trustworthiness, often before we even examine the actual data.

The MIT research team found that persuasive communication operates on multiple levels beyond just data presentation. When encountering a chart on social media, viewers frequently make instant subconscious assessments about who created it based solely on visual style, which then determines whether they find the information credible.

“We showed that visualizations convey more than just the data itself. They also communicate social signals,” explains Arvind Satyanarayan, co-author of the study presented at the IEEE Visualization conference. This represents a newly identified layer of visual communication that plays a significant role in how information is received.

In their experiments, researchers interviewed Tumblr users and conducted broader surveys, showing participants charts both in original form and with text, titles, and axis labels removed. Remarkably, even with blank graphs, respondents formulated detailed assumptions about the creators and their intentions.

Based solely on visual elements like color schemes and fonts, participants made specific judgments, such as hypothesizing that a chart was “created by a conservative boomer” or determining it resembled “a corporate presentation.” This readiness to form opinions based on minimal visual information highlights our vulnerability to misinformation.

“If you’re scrolling through social media and you instantly dismiss a chart as something created by an influencer for attention, that shapes your entire experience—before you even dive into the data,” Satyanarayan notes.

The research identified key design elements that influence these judgments: color palette, font choice, and overall graphics quality. In one striking example, a participant viewed a chart featuring Georgia and Texas flags with a red-and-black color scheme. Despite the absence of any text, they immediately declared, “That looks like something a Texas Republican would put on Twitter.”

This phenomenon transcends education levels and data literacy skills. Graham M. Jones, an MIT anthropologist involved in the research, explains: “It’s not a mistake that people draw these conclusions. It requires a great deal of cultural knowledge about where data visualizations come from, how they are made, and how they are distributed. Drawing conclusions is a feature, not a bug, of how we use signs.”

The researchers even developed a typology of specific associations. Charts considered “too beautiful” triggered suspicions of marketing manipulation, while graphics designed in a simple “default Word settings” style were perceived as honest, albeit amateurish. As one respondent put it, “If something is so pretty that you can tell a graphic designer approved by the marketing department made it, they are trying to sell you something.”

Trust in visual information extends beyond mere aesthetics. The MIT team’s statistical model revealed that the perceived intentions of the author explained 38% of the variance in trust, while design beauty accounted for only 13%. The most influential factors were the perceived goal of the communication (informative versus manipulative), the creator’s presumed competence, and critically, whether recipients felt the author “was on their side.”

This finding has troubling implications for information dissemination in an increasingly polarized society. Michelle Morgenstern, a study co-author, warns that this psychological tendency can fuel misinformation. “If the color scheme and layout of a chart suggest it was created by someone from ‘the other side,’ the recipient might reject even true data,” she explains, adding that “Simply engaging with it would be an act of betrayal against one’s own identity.”

The research coincides with other studies showing that only 15% of the most active social media users are responsible for spreading 30-40% of false information online, and approximately 14% of U.S. adults admit to knowingly sharing false political content.

While the MIT team doesn’t offer simple solutions to this complex problem, they emphasize that ignoring this invisible layer of communication represents a significant risk, particularly in scientific communication. As visual elements increasingly shape our information landscape, understanding these psychological mechanisms becomes crucial in combating the spread of disinformation.