Naked Statistics by Charles Wheelan

Fear of too much information has taken a bludgeoning but survives, discovers Andy Field

February 7, 2013

With every click the Stella McCartney stilettos made as they accelerated down the corridor, Harry’s heart pounded; then it stopped. The thud of Professor Sam Burton’s hands on the swing doors of the office defibrillated it back into life. Burton was ruthlessly precise, levering the doors open enough to allow her through but with sufficient backdraft to float her Alexander McQueen jacket away from her slender frame. She liked to show off her slender frame and doors were a useful tool in this pursuit. She was too attractive, well dressed and young to be a professor, and Harry was too chiselled, buff and cleanly shaven to have spent the past 96 hours poring over data. Even his bespoke suit from Sir Tom Baker of London showed not a crease, nor a Pot Noodle stain. Data analysis had never been so sexy.

“What do we have?” asked Burton urgently.

“Even though I’m only 22, my 10 years of forensic and academic experience tell me it’s a book about statistics,” he replied. “It’s a silent witness.”

“What do we know, Harry?”

“We were tipped off last week by the publishing house W.W. Norton that something was going down in Wells Street in London. It said that the dread of statistics had been brutally murdered and to expect wit and sheer fun. Immediately we suspected Freakonomics or Super Crunchers: they fitted the profile. The scene was reassuringly predictable: we found the usual suspects - descriptive statistics, correlation, regression, probability, statistical inference and a core of relevant, engaging examples. The crime scene was also spattered with clarity, wit and humour, but not in the abundance suggested by the tip-off. We’re treating it as attempted murder.”

“Did you consult Peter in New York?”

“Yes. He said they’d had exactly the same tip-off about Fifth Avenue. He sent over their report: it concluded that the murderer was witty, engaging and a thoroughly likeable chap. They suspected Naked Statistics by Charles Wheelan. However, something about their report unsettled us: their crime scene was littered with baseball batting averages and quarterback pass rates. A lot didn’t make sense to us.”

“What? No soccer?”

“None.”

“Not even a picture of David Beckham in his underpants?”

“I’m afraid not.”

“How on earth can we process this information without semi-naked Becks to keep us company? Where does this leave us?”

“At first we thought we were dealing with different killers, but once we analysed the evidence further we found similarities. Some fabulous chapters on probability and abuses of statistics melted our icy British exteriors and we started to thaw towards the style. Although chapters sometimes began with examples that were then not followed up until pages later, we developed enough memory capacity to cope. Once we co-varied for our tendency to think that humour and engaging examples have to equate to the crass and bizarre, we found that our views had started to mirror those of the New York team very closely. We think it could be the same killer.”

“Is there anything missing?”

“Although a considerable arsenal was employed to make the killing less painful, we’re not sure that the dread of statistics is completely dead, but it has certainly taken a bludgeoning. Also, it was a fairly traditional attack: we had hoped to find more criticism of null hypothesis significance testing, coverage of robust methods such as bootstrapping and more prominence to effect sizes.”

“What can we conclude?”

“We have a very well-executed attack on the dread of statistics by Wheelan. He is a likeable assailant, who has employed some clever and interesting examples in his onslaught. His methods may not be to everyone’s taste, but if you embrace rather than resist (my initial urge), then you will be enlightened and entertained in equal measure.”

“Harry, one final question. Why are we pretending to be a TV show in a book review?”

“Because we’re stuck in a pastiche of Wheelan’s frequently used CSI analogy.”

Naked Statistics: Stripping the Dread from the Data

By Charles Wheelan

W.W. Norton, 320pp, £18.99

ISBN 9780393071955

Published 7 February 2013

Register to continue

Why register?

  • Registration is free and only takes a moment
  • Once registered, you can read 3 articles a month
  • Sign up for our newsletter
Register
Please Login or Register to read this article.

Sponsored