Misuse of statistics in political arguments is not easy to regulate. Mostly the statistics referred to do exist, although their provenance may be hard to discern. And they do relate to the policy under discussion, even if that connection is not clearly defined. But that does not mean anything goes: even when accurate statistics are presented, people can take the wrong impression. Caveats are important to have, but footnotes rarely make the headlines and statistical ideas are sometimes very counterintuitive.
So what does it mean to be misleading, and how much authority should statisticians have over what is said? The Office for Statistics Regulation have been wrestling with this problem for long enough they have now published their view. Join our expert panel chaired by Timandra Harkness to discuss one of the more challenging issues arising from greater use of statistics in public life.
All welcome
This is an in-person only event
(A recording should be available post-event)
Statistics have great potential to inform, to make sense of our world. They can also persuade people of the importance of social issues. However, this power is open to abuse by various routes, whether distraction misrepresentation, obfuscation or distorted comparison. Sifting political debate from misuse has been on the minds of people at the
Office for Statistics Regulation for some time and they have gone as far as to consult a philosopher on what it means to be misleading. Their
thinkpiece was originally published in 2020, and developed with some advice from philosopher Jenny Saul. Since then they have consulted with various groups and published their
reflections in 2021, but this will be the first time they have a public discussion.
In an epoch of concern about fake news and misinformation, there is a tremendous amount of focus on how to combat misuse of scientific evidence of all kinds, including official statistics. But how should we identify genuinely misleading uses of statistics, and how should we confront misleadingness when it arises? Such sins of commission can be contrasted with omissions, where suitable data is not used, but sometimes it is not. And then there are situations when the statistics are simply misunderstood, where the complexity, communication and public understanding are tested. Lots of people see dishonesty in political statements, but it is often more complicated than that. And on top of that, who are statisticians to tell people what they can and cannot say in a democracy?
More use of evidence in public policy ought to be a good thing, but perhaps not if it is cherry-picked, misrepresented, or misconstrued. What sort of role ought statistics to play in public life, and is anyone really any more trustworthy than their last claim? Join us not to answer the question, but to discuss the terms of engagement, and what statisticians can add. The event will include a summary of the pragmatic approach taken by the UK’s Office for Statistics Regulation (OSR) from Ed Humpherson. The panel will be chaired by Timandra Harkness, who will bring to bear her expertise on how to disagree.
Anna Powell-Smith
Director, Centre for Public Data
Ed Humpherson
Director General, Office for Statistics Regulation
Paul Kiff
Managing Director, Research Advice Service
Michael Blastland
Freelance broadcaster and journalist
Chair
Timandra Harkness
Author and Broadcaster
Tom King for RSS Data Ethics & Governance section
FREE to RSS Fellows
£10 for non-Fellows