How do stories shape what people believe, and how do those stories spread?

This project tackled an increasingly urgent challenge: understanding how narratives—especially those shared in news outlets and on social media—spread through our society and influence public opinion. While much of the public conversation has focused on what stories are true or false, we set out to explore how stories are structured and how that structure affects their spread.

At the heart of our work is the idea that stories are not just collections of facts—they are shaped by causal relationships that help people make sense of events. Some narratives are more compelling because they provide strong explanatory links between causes and consequences. Others are harder to follow or feel incomplete. We hypothesized that this "narrative structure" plays a key role in shaping whether a story catches on, independent of its content or truth value.

To test this idea, we developed a new model of narrative diffusion—a way to measure how compelling a story is based on its internal logic and how that influences the way people adopt and share it. We conducted experiments with hundreds of participants, showing that narrative structure alone can explain nearly 70% of the variation in how widely a story spreads. In other words, stories that are better connected—where each event builds clearly on the next—are far more likely to influence people.

We also discovered that people vary in how strongly they respond to narrative coherence. Some individuals are highly sensitive to the internal logic of a story, while others are more influenced by what people around them are saying or doing. Interestingly, we found that early adopters of a narrative tend to rely more on the structure of the story, while later adopters are more likely to be swayed by social factors, such as peer influence or similarity to others.

Beyond the lab, we tested our model using real-world news articles and found that it could accurately predict which stories were more likely to be shared online. We are now working on tools that use artificial intelligence to automatically detect narrative structure in large datasets—an effort that could help journalists, educators, and fact-checkers better understand how stories take hold.

Why this matters

In today’s digital world, stories compete for attention at massive scale—and not all of them are told in good faith. Understanding how narratives spread helps us move beyond slogans like “fake news” and toward a deeper understanding of how public beliefs form, evolve, and solidify.

Our work shows that the structure of a story can be just as important as its content, and that different people respond to different cues when evaluating what to believe. These insights have important implications for how we engage with information online, how we design public health or civic campaigns, and how we respond to misinformation.

Ultimately, this project contributes to a more informed and resilient public sphere. By identifying the hidden dynamics of narrative spread, we can begin to design better strategies to support healthy democratic discourse—where people are not just told what to think, but are equipped to evaluate how and why stories shape the world around them.

Last Modified: 05/07/2025
Modified by: Joshua Introne