Richard Feynman, the Nobel prize-winning physicist, described scientific knowledge as "a body of statements of varying degrees of certainty - some most unsure, some nearly sure, none absolutely certain".
His words resonate in an age when publicising results is a requisite for grants and researchers are encouraged to engage with the public. This places an extra burden on them to emphasise the uncertainties inherent in their work.
At the British Science Festival in Surrey last month, scientists debated how to tackle this thorny issue. Many speakers suggested that explaining uncertainty to the wider world is daunting but not impossible.
Although science's uncertainties are often very small compared with those in everyday life, people judge scientists harshly for them, said Adam Corner, research associate in psychology at Cardiff University.
In research published in the Journal of Experimental Psychology: Applied, he found that members of the public discriminated in their approach to uncertainties in scientific scenarios, such as the probability of a volcano erupting, compared with more mundane events, such as the date of a social event.
Unsettlingly, however, he found that scientists' uncertainty led to their arguments being perceived as weak, and even made the researchers appear unreliable.
Lorraine Whitmarsh, lecturer in environmental psychology at Cardiff, believes that this may be a consequence of how science is taught in schools. Knowledge is presented as fact, so disagreements on the details of science disproportionately undermine people's confidence in them. She argued that schools should do more to teach the "inevitability of uncertainty" in science.
Dr Corner agreed, adding that the public needs to understand how the scientific method operates. This will resonate with many in the academy: a recent Sense About Science survey reveals that only 32 per cent of researchers believe that the public understands the concept of peer review, a cornerstone of science.
But what is the best way to explain risk and uncertainty, especially for researchers dealing with journalists or speaking directly to the public?
The Science Media Centre offers a guide on "communicating uncertainty in a soundbite", including helpful examples to use when discussing why there are disagreements in science or why scientists don't always have the answers.
Dr Whitmarsh advocated descriptive language - such as referring to an event as "very likely" - as opposed to using pure statistics, and said that presenting uncertainties graphically also helps. Where possible, she said, one should "make it real and relate it to experience".
Dr Corner said people respond when scientists "contextualise with a comparison" by using "a non-scientific anchor" they understand.
So relating risk to how likely you are to be struck by lightning or hit by a bus may be useful. But with about 70 per cent of the UK's adult population still regularly playing the National Lottery, some may argue that the public does not understand probability.
Dr Whitmarsh said that the way people assess risk is complex because it requires them to evaluate the likelihood of different outcomes in the context of their consequences.
Some factors, however, are outside academics' control. In a recent survey for the Tyndall Centre for Climate Change Research, Dr Whitmarsh found that the proportion of people who believe that the impact of human activity on climate change is exaggerated has risen steeply in the past five years, despite the scientific consensus rapidly going in the opposite direction.
She said that this may arise from a disparity between people's short- term experiences and long-term predictions - for example, comparing today's weather with climate-change models - as well as self-interest.
People are reluctant to change their lifestyles, she said, so are more likely to latch on to denials than accept an inconvenient truth.
Perhaps the biggest hurdle facing researchers is getting the public to value scientific uncertainty and appreciate that it is born of a desire not to make claims unsupported by the evidence. It is "used as a stick to beat science", Dr Corner said, when in fact it is "a virtue that distinguishes it from ideology".
In the festival's closing session, Lord Winston, professor of science and society at Imperial College London, reiterated the need for public engagement, but pointed out that understanding uncertainty requires "sophisticated thinking".
When struggling to communicate with the public, researchers can at least console themselves that they are not alone.
Tips to help the public understand your research
- Compare risks with real-life examples, and put probabilities in the context of outcomes
- Use descriptions, not just numbers, and employ graphics where appropriate
- Take time to explain the sources of uncertainty and the importance of not glossing over them.