---

Keywords: Statistics, data analysis Title: The Art of Statistics: Learning from Data Author: David Spiegelhalter Publisher: Pelican ISBN: 978-0241398630 Media: Book Verdict: Heartily recommended for anyone with even a vague interest in statistics

Buy US Buy UK

David Spiegelhalter is probably the most prominent statistician in the UK — in particular he is the go to voice the media turn to when some new extreme claim is made about health risks. And it's with good reason — he can get to the heart of the matter and make things intelligible to the lay person without dumbing down or veering into technicalities that obscure rather than elucidate. This is a rare talent and one that promises much in terms of his new book The Art of Statistics — Learning from Data.

The first thing to say is that the book does not disappoint. It is a joy to read — it manages to be engaging throughout, even when tackling some of the more conceptually challenging topics. For those with a fear of maths, and there are plenty out there who glaze over at the thought of complex formulae, this is not a book that will strike fear in your hearts. But this is not to say that the book is in anyway simplistic or patronising. Instead Spiegelhalter manages to explain the concepts, even tricky areas such as statistical inference, with well thought-out examples, problems and plenty of illustrative examples from the news. It might be light on formalism, but it?s not light in terms of content.

There are of course recurring themes — and possibly the most important is the idea that statistics is a tool that is used to understand the world. It's not a toolbox of instant answers — used correctly stats reveals unexpected findings and can uncover hidden relationships buried in data. But of course in practice stats can be used to obfuscate or to intimidate. For academics there are temptations to skew results to generate headline grabbing items or to generate significant results that ensure publication and citations.

There is even some discussion of the statistics wars — with different ideological or philosophical schools disputing the very grounding of their discipline. Spiegelhater is a leading Bayesian, and he explicitly nails his colours to the mast, but in a very even-handed way. There is even some discussion of the historical roots of these disputes, which makes for interesting reading for those of us on the outside looking in.

While stats students will find a lot to enjoy here, this isn't really a book designed to teach people how to do statistics. It's a not a text book ? though it does make me wish that he would write one, with his skills in both analysis and writing it could really be something great. No, this is a book that is designed to give the general reader an understanding of what it is statisticians do and how they do it. It's a book that is well-named, and the art side of statistics and data analysis shines through. The examples, such as the taking apart of the apparently massively increased risks of cancer associated with eating bacon, do the job perfectly.

Overall, this is a book that is heartily recommended for anyone with even a vague interest in statistics. And given that the media is constantly awash with claim and counter-claim, nearly always backed up with doubtful numbers, that ought to be pretty much every individual in the country.