Events | 2009.12.18
Bad Science: A vintage year. Expect more in 2010
It has been a vintage year for dodgy science in government. We saw reports on cocaine that were disappeared, dodgy evidence to justify DNA retention, and government advisers who estimated the cost of piracy at 10% of GDP, to media applause, and then failed to tell everyone they\'d got the figure wrong by 1,000%.
There were fantasies from the security services of using mass surveillance to spot terrorists from their communication patterns, although the basic maths of screening predicts a crippling rate of false positives when trawling for such rare outcomes.
And there was the government\'s claim to have captured £50m of heroin in Afghanistan which would "starve the Taliban of funding": in reality the haul was worth £100,000, while last year the export value of opiates at border prices with neighbouring countries for Afghan traffickers was roughly £2bn.
A £6m Home Office drugs education study was published with no results, because it was so flawed it couldn\'t produce any; we saw MPs being foolish about cervical screening and moon magic, and then when they didn\'t like the scientific evidence they got from Professor David Nutt, they sacked him.
If politicians want us to take them seriously on the evidence for global warming, they have to show they care about evidence everywhere.
We saw the benefits of Tamiflu overstated in the Parmageddon coverage but also uncovered more windows into how evidence is distorted. Industry-funded studies are massively more likely to get into respected academic journals than government-funded studies, even when there is no difference in methodological rigour and quality: all that lovely advertising revenue, perhaps. In Australia Elsevier produced a whole pretend academic journal just for Merck.
On the regulatory front, we discovered that despite trial registration being supposedly compulsory, a quarter of the trials in the world\'s most important journals still aren\'t registered, and the MHRA took 21 months to change the side-effects labels on statins, because one drug company objected (I\'ll find out which one by next year). The only good news is that the industry has failed to stop Indian companies making cheap copies of Aids drugs for people in developing countries.
In further Aids news, Christine Maggiore, poster person for the success of refusing medication, died of pneumonia, and we saw denialism promoted in an Elsevier academic journal (now retracted) and in a foolish feature film, shown by (although they pulled it) and promoted in – of all places – the Spectator.
Elsewhere, we saw that exercise makes you fat, coffee makes you see dead people, and Facebook causes cancer, while housework prevents it, in women.
There was industry-standard front-page wrongness about vaccines (and the Irish Daily Mail campaigning for the cervical cancer vaccine, while the UK Daily Mail campaigned against it). We saw a man in a coma communicating with a method shown not to help people communicate, hideous distortion of research on rape, and much more, although we also found that around half of all academic press releases fail to flag up studies\' flaws.
In libel news, Peter Wilmshurst is being sued for criticising the results of a cardiology trial he had worked on, Simon Singh is being sued by the British Chiropractic Association, but now everyone knows how dodgy their claims are, and the Guardian got £365,000 of £535,000 costs successfully defending a libel case from Matthias Rath, which means the cost of winning is slightly less than the average cost of a home in the UK.
Lastly, lawyers from LBC 97.3FM threatened me with copyright law for posting a foolish anti-MMR broadcast, and as a result, the thing they wanted to disappear ended up being discussed on 160 websites, an early day motion, newspaper pieces and ITV news. There are a lot of people out there who want people like us to shut up. That\'s their bad luck. See you in 2010.
For more information, please visit
http://www.guardian.co.uk/commentisfree/[...]-science-goldacre-vintage-year
Instagram is adding automatic in-app text translation - The Verge views: 1530
Here’s a ‘Brand-New’ Massive Multilingual Dataset for Machine Translation views: 1527