Bad Science by Ben Goldacre

Just to clarify things, this is a book about exposing bad science, not that Dr Goldacre is a bad scientist. And as there is only one of him, it would be difficult to construct a suitable double-blind placebo-controlled randomised trial to find out whether he is… anyway, I digress. I’ve just finished reading the aforementioned book, and can highly recommend it.

The book can probably be summarised with one quotation:

The most important take home message with diet and health is that anyone who ever expresses anything with certainty is basically wrong, because the evidence for cause and effect in this area is almost always weak and circumstantial

I would argue that this can be generalised further. Anyone who expresses anything with certainty is wrong, because the truth is almost always more complicated, far from certain, and subject to a multitude of caveats.

This book leads you through the process of science, with some amusing, and many scary anecdotes. Scary, in that it is difficult to comprehend that people think and act in the ways described.

As with a lot of popular science books, personally, I would have liked Ben to have gone a little deeper into the maths and statistics[1]Specifically, I think that some further discussion of statistical significance would have been interesting. Especially from someone like Ben who has a good working knowledge of the topic, but isn’t … Continue reading. And judging by the tone of the book, so might he. However, I appreciate that his publishers probably disagree — thinking that mathematical formulae scare people. Maybe they do — but I want to argue for them.

The concluding chapters go into the detail of the MRSAscare and the MMR hoax, and the misleading, biased, and in cases false media coverage thereof. What is written in newspapers is not necessarily the truth, and certainly not if it is in the “comment” section. Reading this book will also make me question the “experts” quoted in the media, especially where there are no references, evidence, or links to back up their opinion. One must remember, that it is only their opinion, unless they can back up their claims with evidence.

If this doesn’t convince you that you need to read this book, there is a chapter avaiable to read — for free — on Ben’s badscience.net website which gives a flavour, although perhaps not the best introduction, to the book.

Notes

Notes
1 Specifically, I think that some further discussion of statistical significance would have been interesting. Especially from someone like Ben who has a good working knowledge of the topic, but isn’t a pure statistician