A fascinating keynote talk, 'Statistical literacy and the media' was this year's Significance lecture, given by David Chivers, professor of economics at Durham University, and his cousin Tom Chivers, science editor at UnHerd. Together, they have recently published a book, How To Read Numbers - which sheds light on the common misinterpretations of statistics in the media, and makes suggestions of how to do better.
David started off the session highlighting the challenge of properly understanding the term 'statistical significnance'. It's not just journalists who are guilty of this, but researchers themselves focus on hunting for 'statistically significant' results rather than focusing on true effect sizes, resulting in 'p-hacking'.
Tom, who's won two RSS awards for statistics in journalism, gave us his take on why numbers go wrong in the media and presented his Statistical Style Guide - aimed at preventing the miscommunication of statistics. The guide helps put numbers into context, something that is often disregarded by the media.
After a rich Q&A, it was concluded that we all play our part in correctly communicating statistics: from academics, to journalists, to governments and the education system.
More on this topic can be found in David and Tom’s brand new book, 'How to Read Numbers'.