Take a photo of a barcode or cover
A review by edent_
How to Make the World Add Up: Ten Rules for Thinking Differently About Numbers by Tim Harford
5.0
This is a lovely and useful book. It contains ten simple rules for making sense of statistics. Every day we're bombarded with hundreds of seemingly contradictory conclusions from an array of confusing statistics. How do we sort the bullish claims from the bullshit?
I was particularly impressed with the book's full throated endorsement of open data and open source. If statistics aren't public - they can't be criticised. If their methodology is secret - they can't be examined. Both are needed for a healthy statistical debate.
The thing which struck me the most was just how easy the rules are to follow. And then I was swiftly batted around the head by the realisation that I only tend to apply them to statistics I'm predisposed to disagree with. All humans are fallible - and this book is a constant reminder that we all need to up our game.
Two very mild points of criticism.
There isn't much new here. If you've been following along with the statistic discourse, or are a regular reader/listener of Tim's work. But it put together well, and provides a great overview of why we trust statistics as well as how to trust them.
Many of the statistics and anecdotes are about America. I appreciate that their hegemony looms large in our legend - but a few more stories from outside their sphere of influence would have been interesting. Perhaps, unsurprisingly, the book is being released in the USA and Canada as "The Data Detective".
Thanks to NetGalley for the review copy. The book is available in all good bookshops (and, statistically, a few bad ones) now.
I was particularly impressed with the book's full throated endorsement of open data and open source. If statistics aren't public - they can't be criticised. If their methodology is secret - they can't be examined. Both are needed for a healthy statistical debate.
The thing which struck me the most was just how easy the rules are to follow. And then I was swiftly batted around the head by the realisation that I only tend to apply them to statistics I'm predisposed to disagree with. All humans are fallible - and this book is a constant reminder that we all need to up our game.
Two very mild points of criticism.
There isn't much new here. If you've been following along with the statistic discourse, or are a regular reader/listener of Tim's work. But it put together well, and provides a great overview of why we trust statistics as well as how to trust them.
Many of the statistics and anecdotes are about America. I appreciate that their hegemony looms large in our legend - but a few more stories from outside their sphere of influence would have been interesting. Perhaps, unsurprisingly, the book is being released in the USA and Canada as "The Data Detective".
Thanks to NetGalley for the review copy. The book is available in all good bookshops (and, statistically, a few bad ones) now.