Fascinating new book that shows how easily we’re misled by statistics

Average number of legs for Swedes?Less than two! Just one of the eye-popping facts in a fascinating new book that shows how easily we’re misled by statistics

  • Anthony Reuben is a journalist who was made the BBC’s first head of statistics 
  • He had some catching up to do and now he’s helping the rest of us catch up, too
  • Reuben can now spot a duff statistic from 100 yards (or possibly 91.44 metres) 

Popular science 

Statistical : Ten Easy Ways To Avoid Being Misled By Numbers

by Anthony Reuben (Constable £14.99, 240 pp)

Statistical : Ten Easy Ways To Avoid Being Misled By Numbers by Anthony Reuben (Constable £14.99, 240 pp)

Oh, it’s a funny old thing, but whereas it’s not socially acceptable to admit that you can’t read or write, it’s perfectly reasonable in most sectors of British society to admit that you are a duffer with numbers.

‘Oh yes,’ says everyone with pride, ‘I was terrible at maths at school. No idea about it at all!’

Why is this? It’s not as though numbers don’t matter. For better or for worse, this is now a numerical world. Computer nerds rule the internet. Football clubs are run by accountants. It doesn’t matter how good a film is, it matters how much money it takes.

A book like Anthony Reuben’s is therefore timely. I have read a few like this before, trying to help the innumerate through the vast impenetrable thicket of mathematics, and the problem with most of them is that they are written by mathematicians.

I yield nothing in my admiration for professional mathematicians, but for most of them, maths is dead easy. They don’t understand why you don’t understand it.

Reuben, by contrast, is a journalist who, for reasons too complicated to go into here, was made the BBC’s first head of statistics. He had some serious catching up to do. Now he’s helping the rest of us catch up, too.

Reuben can now spot a duff statistic from 100 yards (or possibly 91.44 metres). One of his favourites was the claim, published in a leading newspaper, that ‘a Saturday night in costs hosts up to £118.29 on average’.

‘Notice,’ says Reuben, ‘how the writer has combined “up to” with “on average” to give a completely meaningless figure.’

The figure is based on the idea that people are inviting four guests over to watch Strictly Come Dancing or The X Factor, and buying refreshments for them. The spending per person is £11.24 on alcohol, £10.92 on takeaways, £6.23 on snacks (seems high?) and £6.32 on soft drinks.

That takeaway figure turns out to be based on ‘desk research’, which involved finding out how much a set menu for four cost at Chinese restaurants in Cardiff, London and Manchester, and Indian restaurants in Fife, Nottingham and Bournemouth, and averaging them out.

He saves the best until last. The survey established that 55 per cent of women bought a new outfit to wear in front of the television, spending up to £100. In other words, lots of spurious made-up statistics are combined into one huge, daft, made-up statistic. ‘It’s a triumphantly terrible piece of work,’ says Reuben.

Surveys, he says, are often inherently dishonest and not to be trusted. One hilarious example he gives is from a press release which says 30 per cent of people would consider taking a holiday at a site affected by radiation.

No description of who these 30 per cent are, how the firm reached that figure, who conducted the research — nothing.

The press release was sent by a company which makes equipment that protects you from radiation. Holidays in Chernobyl, it seems, are flying off the shelves. Perfect for the hyper-fast tan.

Big-number costings are another of his targets. Newspapers are constantly reporting them as gospel. People not having enough sleep costs the economy £37 billion a year — the same amount poor customer service is costing firms.

Lack of neighbourliness costs the economy £14 billion a year, financial crime costs us £52 billion (even though hardly any of it is ever reported), and absenteeism costs us £100 billion a year (a suspiciously round number).

Muscle and joint pain costs the European economy €240 billion a year, and careless Brits have lost more than £3.2 billion worth of wedding and engagement rings over the past five years. Or maybe they didn’t.

Journalists, it turns out, are as useless with numbers as the rest of us. Facts they check, but statistics they often swallow whole.

Number-crunching: Stan Laurel and Oliver Hardy in A Chump At Oxford in 1940

Reuben is admirably sound on more prosaic mathematical problem areas, such as averages and percentages, which surprisingly few people seem to understand.

Clarity is his watchword: he explains dozens of difficult concepts and makes them seem straightforward. His chapter on opinion polls, and why they’re often wrong, should be a set text at journalism schools. And there’s a lovely humorous undercurrent to it all.

In October 2012, a French woman received a phone bill for €11,721,000,000,000,000. She phoned her provider and suggested it might have made a mistake. It said it was definitely correct and offered to let her pay in instalments. She worked out that if she had paid in instalments equal to the entire output of the French economy, it would have taken her 6,000 years to pay the bill. The phone company later admitted that the bill should have been €117.21.

Reuben quotes approvingly a Swedish statistician, who was the first person to point out that the average number of legs for people in Sweden is fewer than two.

No one has more than two, and a small number of people have fewer than two, so the average is very slightly below two. What this means is that almost everybody in Sweden, and indeed almost everyone in the world, has an above-average number of legs.

As Disraeli once said: ‘There are three kinds of lies: lies, damned lies, and statistics.’ All statisticians, says Reuben, hate this quote.

Source: Read Full Article