Counteracting Narratives: Evidence from An Online Experiment

The research article by Manwei Liu and Sili Zhang is now published in The Economic Journal!

We asked economist Sili Zhang to share their key insights.

Discover how exposure to biased narratives influences information processing and belief formation, even when awareness is high.

Your recent paper explores how people respond to biased narratives. Can you explain what are you examining?

Indeed. It is becoming increasingly clear that people can be influenced by narratives, not just mere facts. In the face of challenges of biased narratives posed to society, many advocate for awareness campaigns, assuming that, once individuals recognize the use of biased narratives, they will be able to safeguard themselves. Our study looks at whether people can "undo" the influence of biased narratives once they’ve been exposed to them — even when they know those narratives are slanted. We are interested in knowing: Is awareness enough?

This is a highly valuable skill nowadays, as we frequently encounter narratives about the same event from polarized perspectives. It’s a significant challenge to remain aware that information should be validated from multiple viewpoints, and the first opinion we come across is not always the correct one. How did you test your idea?

We ran an online experiment using a real but less politically charged topic: genetically modified (GMO) mosquitoes used to control disease. This topic was carefully chosen for its relative obscurity to most people, in order to ensure that pre-existing strong opinions would play a minimal role, providing the best chance to counteract the narratives. We gave people one of three things to read — a pro-GMO narrative, an anti-GMO narrative, or both. Importantly, the two different narratives were crafted based on the same set of facts and we told everyone clearly that the stories were biased as such and assigned randomly.

Then, we gave participants the option to read more arguments from both sides and measured how their attitudes changed.

And what did you find? How do people react when confronted with opinions that contradict themselves and what does this do to their own opinion?

We find that even when people knew they were reading a biased story, it still affected their opinion. People who read the pro-GMO story became more supportive of GMO mosquitoes, and vice versa for the anti-GMO story — compared to those who read both.

Even when we gave them the chance to read balanced arguments later, their opinions didn’t shift much. This is the case even when the initial exposure happened two weeks ago. So just being aware of bias wasn’t enough to "correct" their initial impressions.

Did you find any way to actually help people overcome the bias?

Yes, and this is key. When we gave people the exact opposite of the narrative they originally read — not just short arguments, but a full narrative from the other side — they were able to update their opinions much more. This suggests that simply giving “more facts” may not be enough; people need to be shown the full story from both sides early on.

What does this mean for media, education, or policy?

It means we can’t rely on just “raising awareness” of bias to fix misinformation or polarization. Once a biased narrative takes root, it’s hard to undo. This supports the idea that exposure to balanced perspectives from the start is crucial. Waiting to correct things later might be too late, but prevention can go a long way.

Read the article: Manwei Liu and Sili Zhang: Counteracting Narratives: Evidence from An Online Experiment In: The Economic Journal 2025

Read as well: https://researchingmisunderstandings.substack.com/p/is-awareness-enough-to-safeguard