In June 2018, Anthony Masters was selected to be a statistical ambassador for the Royal Statistical Society. Here he describes what work he's undertaken in the role so far.
The misunderstanding of statistics can have dire consequences, such as: people under-estimating savings needed for retirement, the mis-evaluation of assets during the 2008 financial crisis, and imprisonments based on faulty interpretation of probabilities.
The Statistical Ambassador programme exists so more people can talk to policymakers, researchers, journalists and the public about statistics. We received media training and had professional photos taken. This article is about what I have done in my first year as a Statistical Ambassador.
The moon landings and lotteries
ITV News claimed that a recent survey suggested that '52% of Brits think the Moon landings from 1969–1972 were faked'. Independently, I investigated this Atomik Research survey result: finding that it was based on an unweighted internet panel sample — with the question using check-boxes. With the assistance of the Market Research Society, I published the data tables for that survey.
The failure to tick a checkbox on a question about what 'do you believe in' was misinterpreted to mean people believe they were faked. Context effects and question format differences are more plausible for explaining the difference with other survey results.
Following a media request, I was quoted by BBC Reality Check — giving estimated probabilities of mortality from various causes and other comparative probabilities. This was based on a ‘crude’ mortality rate from US mortality data in 2015. In hindsight, it would have been better to give annual rates.
An original article and fact-checking SOTY
I wrote an original article for The Spectator website, pointing out serious flaws in the claim that it was 'very likely' that Vote Leave’s overspending caused 800,000 voters to switch their choice in the 2016 EU referendum.
The article had to be written quickly. On reflection, I would have written it a little differently — to give clarity to my reasoning.
Near the end of 2018, the statistical ambassadors were asked by the Royal Statistical Society to help fact-check entries to their ‘Statistics of the Year’ competition. Four of us looked at all the shortlisted submissions, giving rigorous ratings for each one.
Blogging, posters and judging
I regularly write articles on Medium about statistical matters. Some of these Medium articles have been shared widely. My article on UK employment trends was shared by the UK’s Deputy National Statistician.
My article about whether the British public wishes to leave the European Union without any deal was quoted by Full Fact — highlighting acquiescence bias in how people respond to agree-disagree questions.
Using R, I drew and analysed data on how British MPs use the social network Twitter. This poster was then presented at the Conference on Applied Statistics in Ireland.
I was also a judge for the 2019 RSS Journalism awards. The categories I looked at were Explaining the Facts and Data Visualisation.
In pursuit of corrections
There was an error in a Fathers4Justice press release. Multiple newspapers repeated the false claim that a ComRes survey suggested that four in ten fathers had not seen their children on Father’s Day (or will not see their children next Father’s Day).
That figure (37%) was the estimated proportion of fathers who had not seen their children on Father’s Day or knew someone who had experienced that.