Freya Robb, Senior Press Officer at the Science Media Centre, shares tips from Prof Sir David Spiegelhalter’s Stempra 2026 Communications Training Day session, reminding us that science communication is about making sure the numbers mean something as much as it is about getting them right.
We were lucky enough to have Prof Sir David Spiegelhalter wrap up Stempra’s 2026 Communications Training Day with one of his brilliant lectures on how to communicate statistics well. David is Emeritus Professor of Statistics at University of Cambridge, non-executive director to the UK Statistics Authority and chairs the Winton Centre for Risk and Evidence Communication, so was perfectly placed to help a room full of science communicators get this right.
The session started with a quote from Nate Silver’s book, The Signal and the Noise: “The numbers have no way of speaking for themselves. We speak for them. We imbue them with meaning”, highlighting how important our role is in making data meaningful for our audiences. If we get it wrong, poor and misleading figures can spread far and wide…
With that weighty responsibility resting on our shoulders, David gave some key tips and reminders for communicating statistics responsibly.
1) Remember the hierarchy of evidence
Different types of studies tell us different things, from individual case reports up to systematic reviews of multiple randomised controlled trials. Be clear about what evidence can and can’t show.
Language is the first place this plays out in practice. Experimental studies may support causal language (“x causes y”), while observational studies should use more cautious phrasing (“associated with,” “more likely to”).
Bookmark the Academy of Medical Sciences’ Labelling System. It gives a handy way to understand the significance of research and includes explainers for the different types of studies you might come across.
2) Don’t leave people to interpret big numbers themselves
Big numbers can be really incomprehensible, so put them into relatable terms like “cost per person”.
The same number can sound alarming or reassuring depending on how you do this. Think about how your audience might respond emotionally based on the different ways you could frame the figures.
3) Explain the absolute risks
Relative risks can sometimes appear to exaggerate findings. You’ve probably heard about something “doubling” or “tripling” your risk of a scary outcome, but if the baseline risk is low, the increased risk might actually be very small.
Absolute risk gives your audience the fuller picture, and framing it as ‘X in every 100 people’ has been proven as one of the most effective ways to make that land.
For example: 25g of processed meat daily is associated with 19% increased risk (relative risk) of getting bowel cancer. Ordinarily 6 in 100 people would develop bowel cancer. If those 100 people ate a bacon sandwich every day (approx. 25g of processed meat), we would expect to see 7 in 100 cases of bowel cancer. This isn’t as significant an increase as you might expect.
4) Statistical significance vs. ‘practical’ significance
Even if the results are statistically significant, the context is still important. Think, what is the ‘practical’ significance of these findings? For example, in clinical research, does the size of the effect improve symptoms, quality of life, or survival in a meaningful way?
5) Odds and hazard ratios
We often see these terms in research papers and, while valuable to the scientists doing the work, they rarely translate into anything the audience might be able to picture or act on. Where possible, translate them into visual formats to help show what they mean.
Tools like the Real Risk website can help you convert these statistics into formats that are easier for non-specialists to understand.
Although communicating statistics might not sound like the most glamorous part of a PR person’s role, David’s talk was a stark reminder of why it’s so important to do this well. Our audience was left with plenty of food for thought!
If you missed it, any Stempra member can request access to the full slides from this talk, which includes detailed explanations of understanding and calculating the different figures described above.