2020 UPDATE – A more recent and excellent addition to the list is David Spiegelhalter’s The Art of Statistics, and it would be strange not to mention my own brand new book How To Make The World Add Up.
ALSO – I have a list of recommended books for seeing clearly about statistics over at Bookshop (US) (UK).
A friend recently emailed to ask me for books that might help navigate a world full of statistical bullshit. Here are some recommendations.
I can’t think of a better science writer than Ben Goldacre, who burns with righteous mischief. His Bad Science (UK) (US) isn’t always about statistics, but it’s excellent throughout and an essential read for anyone who wants to understand some of the faults of modern health and nutrition journalism. Wonderful book.
Of course you should subscribe to the More or Less podcast, but you could also enjoy The Tiger That Isn’t (UK) (US). This is the unofficial book of the series, written by More or Less founders Andrew Dilnot and Michael Blastland. A highly readable guide to making sense of numbers in the wild.
Also very good – with more US examples – is Stat-Spotting (UK) (US) by Joel Best. Best’s book has given me some of my favourite examples of bad stats, but it currently seems a bit overpriced on Amazon, alas.
The classic of the field is, of course, Darrell Huff’s How To Lie With Statistics (UK) (US). There’s a sad coda that will tarnish your view of Huff; but this is still a terrific book.
Brand new book by the very splendid Evan Davis is called Post Truth (UK) (US) – haven’t yet read much but looks good.
And finally try Naked Statistics (UK) (US) by Charles Wheelan, who with wit and clarity wrote the similarly excellent Naked Economics (UK) (US).
Best, Dilnot, Huff and Wheelan all cover quite similar ground. If I was picking just one of them I’d go for Dilnot for a UK audience and Wheelan in the US.