Half of MPs unable to solve a simple maths question according to Royal Statistical Society numeracy survey

Can you answer the statistical problem that stumped 48of MPs?

Half of MPs are unable to solve a simple probability question, according to a survey designed to test Parliamentarians’ statistical literacy.

The Royal Statistical Society (RSS) put 101 MPs to the test in a maths based survey, carried out by Savanta ComRes.

The results show more needs to be done to ensure the UK’s elected representatives have the statistical skills they need for the job, the RSS said.

It asked the MPs what the probability is of getting two heads if you flip a fair coin twice.

Only 52% of those surveyed gave the correct answer of 25%. A third (33%) said the answer was 50%, while 10% didn’t know. The rest gave other answers.

Labour MPs were the most successful out of the three main parties, with 53% answering correctly compared to 50% of Conservatives and 45% of Scottish National Party MPs.

Politicians in Scotland and Northern Ireland were more successful (66% answered correctly) than those in England (53%) and Wales (24%), while those representing constituencies in the North and midlands performed better (64% and 66% respectively) than those in London (40%) and the South (32%).

MPs’ statistical literacy has improved since the last time the RSS put their skills to the test 10 years ago, when only 40% gave the right answer to the same question.

Stian Westlake, chief executive of the RSS, said: “Statistical skills are vital for good decision-making and effective scrutiny.

“While we’re pleased to see that it looks like MPs’ knowledge in this area has improved, the survey results highlight that more needs to be done to ensure our elected representatives have the statistical skills needed for the job.”

The RSS put another question to the MPs, quizzing them on their knowledge of averages: if you roll a six-sided dice, and you roll a one, three, four, one and six, what is the mean and mode?

Two thirds of MPs (64%) correctly calculated the mean as three (15 divided by five) and 63% correctly stated the mode as one (the most commonly rolled number).

They were also tested on what the RSS called ‘Bayesian statistics’, with a problem based on lateral flow testing.

Those surveyed were told to imagine they had a diagnostic test for a virus, with a false positive rate of one in 1,000 - meaning one out of every 1,000 people without the virus get a positive result.

They were then asked what the probability is if they have the virus, if they took the test and got a positive result.

The options given were 99.9%, 99%, 1%, 0.01%, ‘not enough information to know’, and ‘don’t know’.

Of the politicians surveyed, 16% gave the correct answer that there was not enough information to know.

A message from the editor: