A lot has been said about the communication of mathematical work during the pandemic, especially about what has been presented officially or in the media. Past RSS meetings have specifically engaged with science journalism. But there has been much less discussion about the policy process, the effectiveness of messages or the role of professionals in doing the communication.
Our meeting at the Isaac Newton Institute (INI) in January aimed to look behind the scenes, and hear a bit more about what good practice looks like, and how experienced people have learned over the years. More than 100 people, from academia, government and the private sector, assembled for two days in a series of talks and discussions, including some familiar names like David Spiegelhalter and Timandra Harkness. However, many were more specialised, working behind the scenes in media, government and other professional roles that are thinking about the public value of insights from mathematical and statistical models.
Talking Maths in Public and the Royal Institution Masterclasses already support a high standard outreach and engagement, particularly for schools. But some of that communication concentrates on describing the mathematics, rather than the conclusions from mathematical and statistical analyses and why they might be important. Communication for the public would allow other groups to set the agenda, ask questions and produce their own analysis. And there is a gap around how mathematical ideas are communicated in policy so democratic legitimacy is developed.
All of the recordings of the meeting are available to watch, but there are a few key observations which are worth reiterating:
Presentations from the health and insurance sectors reminded us that applications are developing and becoming more sophisticated. And all of the speakers reiterated the importance of communicating uncertainty and the profound challenge that presented for different audiences. Although heuristics are common, good calibration and domain assumptions are often taken for granted by users.
An initiative of the same institute in 2015 on ‘maths for public policy’
- Although there is enthusiasm for data as at the top of government, dashboards are being adopted, generally without appreciating why the Covid-19 dashboard was so successful.
- The good practice described may be excellent, but there is much less evaluation of the effectiveness of presentations and whether design decisions are really meeting public needs.
- Impartial communication on matters of public interest is very difficult, because everyone has a stake in big policy decisions, so it is better to admit to being affected.
- Other experiences of maths for policy, like climate modelling, have been stuck at the political stage of getting attention, particularly about urgency, for many years.
did not produce a shift in training for mathematicians, even as discussions concluded communication was a significant issue. Our delegates said they enjoyed the meeting, and the high standard of presentations, and that they learned a lot. But we are determined to follow this by acting on some of the deficiencies discussed.
One issue for us is who to involve: industrial mathematicians start within teams with some support for their development, which researchers may not have prioritised beyond a particular project. More pointedly, the private sector model of product and service development includes communication to justify a business case and market the output, unlike much mathematical research.
We are looking at opportunities elsewhere, and possible partnerships, but it was striking that ‘sci comm’ conferences have not taken up the challenge. Perhaps we don’t know about them, or they are a bit anxious about discussing the pandemic (we were too, but that was why we asked Timandra Harkness to chair, and in the end there was a sense of catharsis from the exchange).
So to stimulate delegate (and broader community) thinking, we produced five vignettes of potential follow on activity to elicit responses in the feedback questionnaire. These covered adapting familiar approaches like the RSS Statistical Ambassadors
scheme, and the Royal Society ‘policy primer
’ course for people on prestigious Research Fellowships, and generic training for ECRs.
More speculative were the ideas about working in partnership with UK government, and we think the Government Analysis Function is interested, supported by the Office for Statistics Regulation. And of course a research programme, perhaps within the standard INI scheme of workshops, bringing together people from philosophy, psychology and mathematicians in academia and policy applications.
There were a few practical links made, and concrete reflections: the new Academy of Mathematical Sciences
will have a policy strand as well as knowledge exchange initiatives. People pitched into the work of SPI-M acknowledged that communication training for policy should have been taken up some years earlier, but a simple policy primer document to read about how systems work would have value in a crisis.
Written by Tom King and Kevin McConway