by Matthew Facciani, Yale University; Aleeza Gerstein, University of Manitoba; Kristi Heather Kenyon, University of Winnipeg, and S. Michelle Driedger, University of Manitoba, [This article first appeared in The Conversation, republished with permission]
Earlier this summer, the four of us spent three days in a room with a dozen others whose jobs sound nothing like ours, or like each other’s: a cosmologist, an emergency management scholar, a genetic counselor, a hydrologist, a neuropsychologist, a health risk communication scholar, a cognitive neuroscientist, a theater director, an investigative journalist and a meteorologist. We’re also from four different disciplines: a sociologist, a human rights scholar, an Indigenous community health scholar and a microbial geneticist.
We had convened through CIFAR, a research organization that interdisciplinary groups to work on hard questions, for a workshop on a challenge that none of our fields can solve alone: how to grapple with, and communicate about, uncertainty in research.
To start, we each brought an object that captured our own struggle with representing uncertainty clearly. One of us brought a handheld weather monitor. Another brought a blank flip chart. A third brought a gachapon, the prize from a Japanese capsule-toy vending machine, which you buy without knowing which toy you’ll get. After the show-and-tell exercise, it was obvious we were all circling the same problem.
The instinct for many researchers is to treat uncertainty as a failing, something to paper over before anyone notices. Our conversation suggested the opposite: that uncertainty is everywhere in every field, and it is worth naming out loud rather than hiding. It is not the absence of knowledge or a lack of results – it is a generative space where curiosity, learning and new understanding are possible.
Most of us also agreed that our struggle to articulate uncertainty makes it harder to share our work clearly with the public, and that glossing over it can damage public trust when people later discover the doubt we left out.
Every field is in the uncertainty business
Uncertainty is not a problem to be solved. Navigating uncertainty is the job, and it never fully goes away.
What do researchers mean by uncertainty? Broadly: Everything we don’t yet know about the thing we are trying to describe, and how much that missing knowledge could change the answer.
It could mean that there is a range of possible outcomes – for example, that it might rain or it might not. Weather forecasters collapse a hugely complex atmosphere into a two-dimensional map, knowing that two skilled people will draw different predictions from the same input.
It could mean that information is incomplete or still evolving, like whether an active wildfire will affect a particular community. A wildfire scientist argued that calling fires simply “unpredictable” risks abdicating responsibility. They cannot say exactly when or where a given fire will start, how fast it will move, or what it will destroy, but they can say with near-certainty that fires will happen somewhere. Treating the whole thing as unknowable can become an excuse not to plan.
Uncertainty could also mean that we have modeled something in the lab but don’t know how well that will translate out in the world. Or it could mean that we are not sure how people will respond to information, such as a wildfire evacuation order.
Reducing uncertainty can help to some extent, but local authorities also need to build mitigation strategies under the assumption that a fire is coming.
As another example, genetic counselors communicate uncertainty with families facing a potential diagnosis. They translate what they know about a health risk at a population level into information that individuals can use to make personal decisions they will have to live with.
Even the most quantitative fields run on judgment. A statistician reminded us that the line between a “significant” and a “nonsignificant” result, which conventionally is dictated using a statistical metric called a p-value at a threshold of 0.05, is as much a social convention as a mathematical one.
A p-value reflects the probability that a result found in a study was not due to random chance.
A cosmologist described the 2014 BICEP2 episode, when a team announced that it had glimpsed the universe’s first instant, only for the signal to turn out to be consistent with dust in our own galaxy. Physics demands extraordinary confidence before calling something a discovery, but even that assumes you are measuring the right thing in the first place.
An investigative journalist put it most plainly: “Beyond a reasonable doubt” is not the same as absolute certainty. A wrongful conviction can happen when a jury or judge forgets the difference.
Most people never encounter academic research directly. They encounter it secondhand: through a news story, a news release, a social media post or a conversation with someone they trust. If researchers do not communicate uncertainty clearly, each handoff becomes another opportunity for misunderstanding.
The hard part is the handoff
Qualitative research from March 2026 found that many scientists across disciplines view communicating uncertainty as a form of translation. They see the practice as trying to faithfully convey complex evidence while adapting it to audiences who often expect simple, definitive answers.
Researchers also described a pressure to sound more certain than the evidence warrants. Journalists and editors want a clean, quotable finding, and in a polarized environment, readers can pick up on any admission of doubt and use it as evidence that the whole field is unreliable.
How uncertainty is received depends less on the data than on the delivery and on what the audience already believes. People rarely evaluate findings directly; they absorb them secondhand and lean on identity and trusted messengers. In practice, that means you take your cues from people you see as being on your side or like you – a family doctor, a faith leader, a neighbor, a favorite podcaster – more than from the study itself.
People with different life experiences, perspectives on health and well-being, and histories with government, authorities and researchers interpret information differently and trust different sources. Those who have experienced discrimination in the healthcare system are less likely to trust medical practitioners and less likely to seek their care. People who understand their health conditions within a cultural or spiritual context may eschew pharmaceutical solutions, or may use them in conjunction with other forms of treatment.
Imagine two people reading the same headline about a rare side effect from a new vaccine. One has always had a doctor who explained things patiently, and reads it as reassuring evidence that the system catches problems. The other has been dismissed or mistreated in a clinic before, and reads the same sentence as confirmation that they were not told the whole story the first time. The number is identical. The conclusion is not.
The reassuring finding, across the research represented in the room at the CIFAR workshop, is that candidness about uncertainty generally maintains or builds trust rather than erodes it.
Researchers who are sincere, relatable and genuinely care are effective messengers for communicating uncertain or evolving information. What erodes trust is vagueness, backtracking and the sense that someone is managing you rather than leveling with you.
However, manufacturing uncertainty when the research is sound can serve to reinforce power structures or political narratives. The clearest example is the tobacco industry, which spent decades funding research and publicity designed not to prove cigarettes were safe, but simply to keep the question looking open. That strategy of producing scientific uncertainty delayed regulation for a generation, and the same playbook was later borrowed by other industries, including on climate change.
What you can ask of researchers
If there was a consensus in the workshop room, it was less a formula than a set of guardrails. These came out of scholars talking to each other across disciplines, but they double as a checklist of questions to ask when you’re reading about a new study or public health recommendation, or watching a weather forecast.
First, is the report clear about what the available evidence tells you, and the boundaries of what it doesn’t? Does it separate the data from the policy options?
Second, does the report set expectations and disclose that the information may change over time, and say how and why it might do so?
Third, is it specific? Plain estimates land better than vague verbal hedging. Hedging can sound like “this could be somewhat higher or lower.”
Fourth, is it transparent about who is speaking and any bias related to the tools, the researchers, the institutions or the funding structure?
Uncertainty is not the enemy of good research or public trust. It is the open question that invites the next one – the space where curiosity lives. In an information environment crowded with confident, unreliable voices, researchers’ willingness to say “here is what we don’t know, and here is how we’ll find out” may be the most trustworthy thing that researchers can offer.
Matthew Facciani, Research Scientist, Yale University; Aleeza Gerstein, Assistant Professor in Microbiology and Statistics, University of Manitoba; Kristi Heather Kenyon, Associate Professor of Human Rights, University of Winnipeg, and S. Michelle Driedger, Professor of College of Community and Global Health, University of Manitoba
This article is republished from The Conversation under a Creative Commons license. Read the original article.