Thursday, 13 December 2012

Thinking Statistically – Uri Bram ****

This is a delightful little book (just three chapters) introducing three of the fundamental aspects of statistics that can get us confused: selection bias, edogeneity (effectively missing external factors which are influencing the outcome) and the use of Bayesian statistics, an approach that is very powerful but makes it easy to go astray.
I wouldn’t quite describe this as a popular science book – there are probably rather too many equations – but it is excellent both as providing a bit of understanding for those making use of statistical methods (it’s all too easy to just crank the handle without understanding what you are doing and thereby come up with the wrong results) and as  an introduction for the general reader who isn’t put off by a little bit of jargon and equations in what is, nonetheless, a very readable little book.
Thinking Statistically is short enough to read in a couple of hours, and I think it’s a credit to the author that I thought ‘Oh, really, I wanted more!’ when I got to the end. Uri Bram’s aim is to get the reader taking a more statistical viewpoint. Not necessarily wheeling out the statistical big guns every time you make a decision, but at least being aware of the statistical processes you are undergoing mentally, often unconsciously.
If you would like to know a bit more about statistics, but find the whole business a bit baffling, this is a good place to start.
You may wonder what the cover has to do with statistics. So did I. The simple answer is nothing.
Review by Brian Clegg

No comments:

Post a Comment