This collection of short writing by Ben Goldacre is mostly drawn from his Guardian Column, Bad Science, where he took apart stories in the media that were built on bad science, often repeatedly. I'm really impressed with Goldacre's work and the amount of time and effort he sometimes had to spend to get past obstructive companies (and sometimes journalists), to get to the original research (or contacting the researchers, if it wasn't available) and lay out not only his conclusions in a very readable manner, but also the process of science and why that's so important, as well as very patiently explaining the basics of epidemiology.

This explanation of how science works, with examples of bad science, forms the largest section of the book, but the range is huge, as you'd expect from having eight years of weekly columns to choose from, ranging from Susan Greenfield's continuing refusal to publish any research to back up her claims that computer games are bad for children to devices that supposedly detect bombs in Iraq.

The book is very readable, each article is fairly short and it's a good book for picking up in bits, or (as Goldacre calls it in the introduction), a statistics toilet book. I appreciate that he's moved on to other things now, but I do miss the Bad Science column. Especially after reading this book, I tend to treat any scientific or statistical claim made in the media with some suspicion. I've started to follow the NHS NHS Behind the Headlines blog which does some of the same things (but only for medicine). Of course this doesn't have the profile of a national newspaper but it's something.