Sunday, October 12, 2008
Bad Science by Ben Goldacre: book review
I have been following Goldacre's column in the Guardian and latterly his blog for a couple of years now, since it is usually the best source of sensible information on any news story that touches on science, technology or medicine. I was fearful that the book might have shared the blogs slightly smug and inward-looking style ('we're clever people and we know everything'), but in fact it is well-written, coherent, and engaging, written in a light and chatty style.
There are extended accounts of the bizarre history of some recent media panics (MRSA, MMR, Dore, and fish oil), but more importantly, the science and 'science' of these stories is examined forensically, so that the reader learns to interpret news stories critically: what does "50% reduction" mean in this situation, what's the sample size. This is worthy and important and should (in time) change the way that news media present their accounts (I have already noticed a survey fatigue, where all involved seem happy to accept their spurious basis).
Perhaps the two most interesting chapters, though, are those on the placebo effect and on our perception of risk. I hadn't known, for example, that painkillers work better if they are packaged better and have been advertised, but it is true. The moral and practical implications of trying to deliver Evidence Based Medicine when this sort of placebo effect can dictate success or failure are a challenge. The chapetr on risk demonstrates at length how bad people are at distinguishing between chance events and patterns, between causation, correlation and coincidence, and how unreliable their accounts of their experiences can be, thanks to selection bias. This important factor explains why people sincerely believe things in the absence, or the face, of objective evidence, whether it is the Bridgend suicide 'cluster', electromagnetic sensitivity, or the Loch Ness Monster.
It should be noted that Goldacre does not adopt a hectoring tone: he argues that these are universal, human, traits; he just wishes us to be aware of them so that we can monitor our belief formation. He notes, for example, the tendency of people to use the limited evidence that moderate drinking is better for your health than teetotalism as a justification for their immoderate drinking. This is why factoids like 'red wine is good for you' are so powerful: there is so much contradictory advice out there that, as Paul Simon said, 'a man hears what he wants to hear and disregards the rest' (an observation, incidentally, that is so perceptive and well-expressed that on its own should preserve his reputation for millenia).
The most contentious part of the book deals with the media and how they report science stories; Goldacre tries to explain why nonsense science so often trumps proper science in media coverage. He suggests that the fault lies with humanities background of most journalists, who find the science impenetrable and feel free to choose the wildest and most exciting of the opinions they are offered. Here he may be wrong, insofar as he assumes that science suffers alone. The sad truth is that the media deals badly with all areas of specialist endeavour. An archaeologist told me recently about press coverage of a Neolithic find; it was dated to 3000 BC, 5000 years ago; in print it became 3000 years old. I wasn't surprised: to the non specialist, it was simply 'very old'. There is an interesting question about how far journalists are to blame in not understanding or whether they undertsand adequately but dumb stories down because their readers won't need or want accurate details. This pervades serious newspapers: strange health advice is dished out in the supplements while in the main paper things are more rational. But perhaps we get the news coverage we deserve: if you want to depress yourself, look at the 'most read stories' list on the BBC News pages.
Goldacre believes that all media, and especially serious newspapers, are engaged in a project to educate and inform their readers; but they aren't. They are there to entertain, mainly: hence the celebritisation of news, with the daily updates of Pete Docherty's battle with drugs, and battles with photographers. But even in the old days, there was a strong vein of cynicism and philistinism in journalism: the attitude that the contents didn't need to be true, just true enough.
Nevertheless, the book is enjoyable and inspiring: the way he benourages the reader to engage with the primary sources should be enough to balance the increasing inaccuracy of the media as reliable informants.