The book does an excellent job of explaining where published statistics come from and what meaning should be ascribed to them, with many examples. Explains how data, averages and risk factors can be deceptive or exaggerated for effect, yet shows how to interpret these numbers without throwing them out entirely. ( )

I'm looking forward to reading this book. The way the media, and even people who should know better, abuse numbers so as to make real risk assessment very difficult is discouraging. For example, the American Institute of Cancer Research says we should eliminate eating bacon because doing so increases our risk of colorectal cancer by 21%. That is true on the face of it and would appear startling until you ask what the baseline is. About 45 of 1000 men will get that cancer, or about 5 per 100 men. If every one of those 100 men ate bacon every day, 6 men would get the cancer, an increase of about 20%, yet actual risk -all other risk factors aside remains really quite low at 5%.(This is an example - I'm not sure of the exact numbers) The principle is the same.

I was frosted recently by the news that taking a multi-vitamin every day did nothing to prevent heart disease or increase life-span. Now my crap detector really started going into overtime. Without even reading the study I can suspect some flaws, because there is no way you can do a truly blind epidemiologic study on 160,000 women over 8 years and exclude all the other variables. Can't be done. And since there was no way to predict ahead of time which participants would be more likely to live longer than others and pair them with similar candidates, how in the world could they come to such sweeping conclusions. And this assumes their diets were absolutely equivalent in all other respects. And then to make things worse, they suggest that in order to achieve those benefits you should eat a diet rich in fiber and greens etc. etc. without a shred of evidence that it would make a whit of difference, for precisely the same reasons as the invalidity of the vitamin study.

interesting essay on the abuse of statistics and numbers, written by contributors from the Economist

A lively guide, based on a BBC radio show, to how to figure out when the numbers cited in the news, etc., make sense. I knew a lot of this before, but the authors mix the information with interesting examples; this edition was revised for the US market so it uses a mixture of UK and US information. I enjoyed it, but it probably could have been a little shorter.