Book review: The Art of Statistics: Learning from Data by David Spiegelhalter
Thursday 25th March 2021
You may have seen, heard, and read articles by David Spiegelhalter on television and in the national press over recent months commenting on aspects of the Covid-19 pandemic. He is a British statistician and Winton Professor of the Public Understanding of Risk in the Statistical Laboratory at the University of Cambridge.
His book, The Art of Statistics: Learning from Data, published in 2019, should be “required reading for all politicians, journalists, medics and anyone who tries to influence people (or is influenced) by statistics”, according to Popular Science. Pretty much, therefore, all of us!
In the book, Spiegelhalter argues that statistics has historically been taught poorly with courses focusing predominantly on learning a set of techniques to be applied to specific situations, with little consideration given to understanding why these formulae are being used and to the challenges that arise when using data to try to answer questions. He identifies that the teaching of statistics is changing, driven by the needs of data science and data literacy, and that a more problem-solving approach is developing. His book uses real-world problems to introduce statistical concepts and focuses on conceptual issues rather than technicalities.
Assessing the trustworthiness of statistics that are presented to us – the fact that statistics can both bring insight into the problems we face or be abused to promote incorrect conclusions – is a core component of the book. Interesting questions, including identifying the luckiest survivor from the Titanic and whether the mass-murderer Harold Shipman could have been caught earlier, are discussed in an interesting and non-technical way. The book is an enjoyable read and is accessible to teachers and students of statistics and to anyone who wants to be more informed about the statistics they encounter in their work and everyday life.