How to Read Numbers: A Guide to Statistics in the News (and Knowing When to Trust Them) by Tom Chivers and David Chivers
Tom is the science editor at UnHerd and has twice been awarded a Royal Statistical Society “statistical excellence in journalism” prize, in 2018 and 2020, and was highly commended for the same prize in 2017. Tom was the science writer of the year by the Association of British Science Writers in 2021. In addition, has been shortlisted for the science journalist of the year by both the British Press Awards and the British Journalism Awards.
His cousin David is an assistant professor of economics at Durham University. He was a lecturer at University of Oxford and completed his PhD from the University of Manchester. David is published in academic journals such as Review of Economic Dynamics, Economic Theory and Journal of Economic Behavior and Organization.
Today, more than ever before, we live in a world of sensational headlines from seemingly thousands of websites who lead with data driven reports. So, is it not a surprise in which they misquote the data in which they present. Of course not. So where does one start to understand a baseline of news data and statistics?
Easy statistics….really
Tom and David are providing a very easy to read guide to statistics. Yes it is very true! You certainly do not need Phd math skills. They examine many sensational headlines with various statistics and breakdown where they mislead readers. This is a very refreshing in explaining in plain english technical mathematical concepts.
The news hype cycle
Today’s alarming headlines really need to be analyzed correctly to see through the hype cycle and get down to the facts. Tom and David reveal how to analyze what the headlines and pie charts present in order to find truth. How Not to Be Wrong by Jordan Ellenberg is a worthy companion to this book!
Stretch your mind
Along the way you will relearn the basics of Biased Data Samples, Simpson’s paradox, The Prosecutor’s fallacy, Texas sharpshooter’s fallacy and even Survior Bias. They are also addressing important elements including the amount of anecdotal evidence, data survey results based upon very small numbers, or the ordinary misinformation when articles suggests ‘statistical significance’ to bolster a claim or position. Indeed, many articles cherry pick data to support their arguments, which often fail to provide the entire data sets.
In conclusion, Tom and David actually provide at the close of the book a ‘statistical style guide’ for journalists. I will keep this book within reach as I will need to refresh my understandings to article in the future. Yes, this book is that good!