Statistics: A Very Short Introduction by David J. Hand

Author: Reviewed by Tom King

Statistics: A Very Short Introduction by

David J. Hand. Published by Oxford

University Press. Paperback. October 2008.

ISBN-10: 019923356X. ISBN-13: 978-0199233564.

136pp. 8 figures. 174x111mm. RRP: £7.99.

So you thought you knew what statistics was? Well, you may be right, but David Hand explains what statistics is now, very different from its origins in compendious tables of figures. The Very Short Introductions are both an honour and a burden: the author must bear the responsibility of representing an entire subject to a lay audience in a very slight publication. For mathematics, they chose Tim Gowers, Fields medallist and a teacher of inspirational insight and profound humility. Even for a Royal Statistical Society president (now ex-President), the bar was set very high: starting the book is a little formal and defensive, with definitions and a response to some negative perceptions but we soon settle down into some insightful examples.

In the preface Hand asserts that the dry and dusty image of a discipline without recourse to imagination relies on a dated focus on arithmetical manipulation. Hand takes on the challenge to show the excitement of modern statistics, a bold aim but perhaps the right one when the potential audience is anyone who might be browsing a selection from this series. In common with many books in this series, the writing is rather academic but the promise of needing no mathematics is kept. In fact, Hand gives a rather broader view of statistics than many statisticians would have considered but this reflects the wide modern relevance. One must trust him to be a guide through the modern discipline: this will not be the book most readers expect but they will not be disappointed.

If you have studied statistics or read about them in order to apply them, you may find that this book is in fact a reintroduction to statistics. Hand lucidly shows how modern statistics should not be seen as a collection of discrete tools but very much more. In order to appreciate the applications of statistics, you need to be aware of the world, particularly politically, but the more you look, the more you will find statistics in the background of diverse subjects from cosmology to jurisprudence. I would have liked more discussion of official statistics and such contemporary issues as performance indicators but I suspect anyone reading would like more depth on their own field of work. There does seem to be an omission of any reflection on the public interpretation and criticism of statistics: so often statistics feature in political discussions but when and how are they useful and how do they mislead?

What is astonishing is that the book can cover the whole of modern statistics in such a slim volume. But it does: factorial experiments, missing data, epistemic and aleatory probability, sampling theory, and all that is by chapter four. In meticulous but unhurried style, the discipline unwinds a logarithmic spiral: accelerating into the subject without losing sight of the focus on learning about our world by looking at data. It is a pity the further reading is unable to collect more books in this style: typical books focus on one topic either from a technical or frivolous perspective, a dichotomy which this one transcends. Unencumbered by mathematical details, this is a comprehensive introduction which may merit more than one reading, particularly of later chapters which have a certain conceptual complexity.