How to Make the World Add Up: Ten Rules for Thinking Differently About Numbers

The book didn't appeal to me at first, but I'm glad I gave it a read. The book pleads that it helps to be curious and critical when it comes to statistics. I know I tend to read the big print and believe the numbers without asking more about what's being measured, who wrote it, and what they want to make us feel. It's not technical whatsoever, but rather quite practical. Interesting anecdotes and case studies around the world of statistics, too!

CRANKY'S BOOK GROUP REVIEW Our book group is all about reading out of our comfort zones, and that’s why we selected this book. There were quite a few absentees this evening, and it could be that this book was a step too far out of some of our comfort zones. Those of us who did read it, however, were really impressed. We all agreed that Tim Harford’s motives were impeccable and that this was a very instructive and worthwhile book to read. It was a particular eye-opener for those of us who weren’t ‘numbers people’ but it also had plenty to teach those of us who had more experience with statistics or research. One of us had actually quoted some of the statistics that he criticised in a talk a few years ago! We agreed that Tim Harford had a great conversational style of writing and that he was one of the few scientific experts who could make highly technical information accessible to the general reader. The real world instances that he used to illustrate his points were very well-selected and frequently mind-boggling, revealing instances of marked gender and racial bias in research and in algorithms formed from ‘big data’, for example. He illustrated very clearly why these instances were problematic and what the source of the problem was. Tim Harford had intended for this book to teach us how to recognise when a piece of research was trustworthy. We weren’t so sure whether he managed to achieve this fully, or whether it was actually achievable. Some of us remarked that it was hard to keep his ten points for analysis in their minds; others made the point that the source data we’d need to look at weren’t easily accessible a lot of the time and that even when they were, they might not be easy for the average reader to understand. We all agreed that he went some way in his aim, however, and that people responsible for using statistics in the media, business or politics REALLY need to read this book. We hoped that it would lead to legislation to make it illegal to hide the source data that produced algorithms affecting real-world decisions on our lives: for credit-ratings, reoffending probabilities, or risk data used in social care, for example. The main problem that we had with the book was that it was very repetitive and could have benefited from a good editor. But it’s such a worthwhile read that we recommend that you read it anyway! We gave the book a rating of 7/10.