Why Some Social Media Posts Feel Trustworthy Before We Know Whether They Are

Imagine scrolling through your social media feed. A polished infographic summarizes a new health study. A dramatic photograph accompanies a political claim. An account shares a chart about climate change, complete with percentages, citations, and a confident explanation. Within seconds, you have already formed an impression. Some posts feel trustworthy. Others feel questionable.

What is striking is how quickly this happens.

Before checking who posted the information, before searching for other sources, and before verifying the evidence, we often have a sense of whether something seems believable. That first impression influences whether we keep reading, whether we share the post, and whether we decide the information deserves our attention.

Most of us assume these judgments are based primarily on truth. But from a psychological perspective, another process comes first. Before we decide whether information is true, we often decide whether it looks like the kind of information that is likely to be true.

That distinction has become increasingly important as social media has transformed how we encounter information. Images, infographics, screenshots, and short videos have become central to public communication about science, medicine, politics, and current events. These visual formats are remarkably effective at attracting attention and conveying complex ideas quickly. But they also introduce a new question. What characteristics make a visual social media post appear credible in the first place?

Over the past four years, our research team at the University of Georgia, the University of California, Davis, Northwestern University, and the University of North Carolina at Charlotte has been studying exactly this question. Supported by the U.S. National Science Foundation, we collected more than 4,000 image-based social media posts covering topics ranging from COVID-19 and vaccines to climate change, elections, international conflicts, and public health. More than 5,300 people evaluated these posts, generating over 42,000 credibility judgments. We were interested not in whether the posts were factually correct, but in what makes them appear credible.

The answer, perhaps surprisingly, is not a single feature. Instead, people appear to rely on a handful of psychological shortcuts that usually work well, but can sometimes be misleading. These shortcuts help us navigate an overwhelming information environment, yet they can also be exploited by persuasive misinformation.

When Easy Feels True

Suppose you are given two explanations of the same scientific finding.

The first is filled with technical jargon, long sentences, and complicated figures. The second presents exactly the same evidence but explains it in straightforward language with a clean visual layout.

Most people would probably trust the second explanation more.

Psychologists have spent decades studying this phenomenon under the name processing fluency: the subjective experience of how easy something is to perceive or understand. Fluency turns out to influence a remarkable variety of judgments. Statements written in clear, easy-to-read fonts are judged as more likely to be true than identical statements presented in difficult fonts. Familiar names inspire more confidence than unfamiliar ones. Even repeating a statement several times increases its perceived truthfulness, a phenomenon known as the illusory truth effect.

Our study suggests that social media follows the same psychological principles.

Across thousands of posts, one of the strongest predictors of perceived credibility was simply whether people found the post easy to understand. Clear images, coherent captions, and messages that could be processed effortlessly were consistently judged as more credible than confusing or ambiguous ones.

This finding should not be interpreted as a criticism of clear communication. Quite the opposite. Scientists, physicians, journalists, and educators all strive to communicate clearly because clarity improves understanding. Processing fluency is useful precisely because it often points us toward reliable information.

The problem is that our brains do not stop at understanding. We often interpret the feeling of understanding as evidence that the information itself is trustworthy.

That distinction matters. A beautifully explained misconception can feel more convincing than a poorly explained scientific fact. Clarity makes information easier to process; it does not make it more accurate.

Details Signal Expertise

Imagine reading two posts about the same medical treatment. One says simply that "research shows this treatment works." The other describes the number of participants in the study, explains the biological mechanism, provides numerical estimates, and discusses possible limitations. 

Without checking any references, many readers would instinctively trust the second version. Our participants did exactly that.

Posts containing concrete, detailed, and specific information consistently appeared more credible than posts making vague or generalized claims. This pattern held across very different topics, suggesting that people use specificity as a broad credibility cue rather than one limited to science or politics.

Again, this heuristic usually serves us well. Genuine experts typically possess more detailed knowledge than non-experts, so specificity often correlates with expertise.

But correlation is not certainty. Conspiracy theories, for example, are often extraordinarily detailed. They contain timelines, institutional names, technical terminology, statistics, and elaborate causal explanations. These details can create an impression of careful investigation even when the underlying conclusions are unsupported.

Our minds appear to treat detail as a signal of knowledge. Most of the time, that assumption is reasonable. But detailed information should invite scrutiny rather than replace it.

We Judge the Message by Its Appearance

We often tell ourselves that we judge information based on its content rather than its appearance.

Our findings suggest that the distinction is not so clear. Posts accompanied by aesthetically appealing, professional-looking images were generally perceived as more credible. Formal writing styles also increased credibility, while memes and obviously manipulated images tended to reduce it.

These findings mirror decades of research beyond social media. Professionally designed websites are often considered more trustworthy than poorly designed ones. Consumers judge products differently depending on packaging. Academic journals invest heavily in visual presentation because design affects how readers perceive the information.

This does not imply that attractive design deceives people. Professional organizations often communicate professionally because they possess the resources and expertise to do so. Over time, people learn that visual quality frequently accompanies institutional competence.

The challenge is that visual professionalism has become much easier to imitate. Today, polished graphics, clean layouts, and publication-quality illustrations can be produced quickly and inexpensively. The visual language of expertise is no longer exclusive to experts.

Relatedly, our participants were also sensitive to coherence. Posts seemed more credible when the image and caption reinforced one another. A photograph that naturally illustrated the accompanying text increased credibility; images that felt disconnected or contradictory reduced it. In psychology, this reflects a broader tendency to favor coherent information. When different pieces of a message fit together, the entire message feels more believable.

But coherence is not evidence. A story can be internally consistent while still being completely wrong.

Looking Verifiable

Posts seemed more credible when they appeared to provide enough information for readers to check the claims themselves. References to institutions, identifiable events, contextual background, or supporting information all increased credibility.

It does not mean that readers actually verified the information. Nor does it mean that the cited sources were reliable. Instead, people appeared to reward the possibility of verification. Information looked more trustworthy when it seemed connected to evidence that could, in principle, be examined.

This distinction between looking verifiable and being verified may be one of the most important challenges facing today's information environment. Modern misinformation often imitates the structure of evidence-based communication. It includes citations, institutional logos, scientific terminology, and detailed explanations. These are included not necessarily because these support the underlying claim, but because they signal credibility.

Why This Matters

One lesson from our research is that credibility emerges from the accumulation of multiple signals rather than any single characteristic. Information that is easy to understand, rich in concrete details, professionally presented, internally consistent, and seemingly verifiable is more likely to be judged as credible than information that lacks these qualities.

These cues are part of how people efficiently navigate an information environment that is far too large for careful verification of every claim. In everyday life, they often point us in the right direction because they frequently accompany expertise and reliable communication. At the same time, they can also be deliberately reproduced. As visual communication becomes easier to create and AI tools make polished content increasingly accessible, many of the traditional cues people associate with credibility are becoming easier to imitate.

Our findings offer a framework for understanding how people make credibility judgments in today's visual media environment. They help explain why some posts immediately inspire confidence while others do not, and they identify the psychological cues that shape those first impressions. For researchers, educators, journalists, and anyone interested in improving online information quality, understanding these cues is an important step toward understanding how credibility is formed—and how it can be earned, or manipulated.

Our full study, Large Language Model-Informed Feature Discovery Improves Prediction and Interpretation of Credibility Perceptions of Visual Content, is available as a preprint on arXiv.