Dec 2, 2012

Science is Awesome (as are Ben Goldacre and Mark Henderson)

xkcd Stand back! I'm going to try science
Credit: xkcd

Long-time readers of this blog will know that science advocacy is one of my biggest passions. By that I don’t just mean going “yay, science!” (though I’m obviously a fan of that too); I mean actually explaining why doing science right is important, as well as being a watchdog for sloppy science and drawing attention to its dangers. One of the themes I kept returning to in my reading and on this blog, gender essentialism, is actually a great example of why we need better science. When we look at the data carefully, any attempts to co-opt the language of science to justify structural inequalities fall apart. As many of you know, people like Cordelia Fine, Mark Lieberman, Agustín Fuentes or Deborah Cameron have written about this subject far better and in more detail than I ever could. I love their work not only because they share my values, but because they believe in fighting bad science with better science.

Recently I had the opportunity to attend talks by two fantastic science advocates, Ben Goldacre and Mark Henderson. These particular talks had nothing at all to do with gender, but as you will see below, the same principles of intellectual honesty, rigour and transparency were at the heart of what Henderson and Goldacre had to say.

Bad Science by Ben Goldacre Bad Pharma by Ben Goldacre

You might remember that a few years ago I read Ben Goldacre’s Bad Science; it was one of those books that fill me with evangelism and have me daydreaming of standing on a street corner giving away copies to any poor unsuspecting soul who happens by. Last week’s event was actually focused on his latest book, Bad Pharma, which I haven’t read yet but really hope to find under my tree this Christmas. The book is about the consequences of withholding scientific evidence from doctors and patients to make drugs more marketable, and this was also the subject of his talk.

Goldacre began by saying that he doesn’t believe that the people who work in the pharmaceutical industry are evil – it’s perfectly possible for honest, well-intentioned and committed people to work within bad systems with very damaging results. The focus of his book is on systematic problems, rather than on the ethics of specific individuals. Similarly, he’s not interested in telling people they’re wrong for its own sake; he’s interested in how people are wrong, and in why factually wrong ideas are so persistent. He doesn’t believe people are too stupid to understand science or evidence-based medicine, so clearly the issue is much more complicated than that.

One of the facets of the problem is this fact that big pharmaceutical companies consistently withhold or distort evidence to make their products more appealing, with no legal consequences. There’s also the fact that the architecture of information in medicine is incredibly poor, and there’s an immense lack of ambition on the part of the medical industry when it comes to collecting and disseminating evidence (speaking of which, The Half Life of Facts, which I reviewed last week, was a perfect companion read to these talks).

Publication bias, a persistent problem in all fields of academic research, can have particularly disastrous consequences in medicine. What the term “publication bias” means is that only positive or successful results are published – to return to the familiar topic of gender, only studies that find differences between men and women tend to be published, whereas the ones that don’t are stashed away in a drawer. When it comes to research into the effectiveness of a particular drug, if we only get to see the studies that do show the drug is effective, we’ll obviously develop a very distorted idea of what the drug in question can do. Only the fluke results are put out there, except we don’t even realise they’re flukes because what we’re seeing is an unrepresentative sample of all the research that has been done.

Replicability is an important concept in science for this very reason: to make sure our results can be trusted, we need to get them not just once, but again and again. The problem, however, is that academic journals are biased against publishing replications of previous studies, and that means nobody much bothers doing them. Lately there have been some efforts to start replication-only journals to correct this, but they’re still in the early stages.

Goldacre also mentioned that medical trials are increasingly done on the poor and the homeless in America, and also on the population of developing countries – which raises all sorts of ethical issues. People, usually vulnerable members of society, are paid to be part of the first ever group of humans to be given a drug, and they risk immense suffering and even death because of side-effects and other unexpected problems. All of this could be prevented with better science before we get to the trial phase. And once we do get to this phase, we need to make all trial results publicly available to prevent future issues. All data needs to be brought together so we can conduct systematic reviews and meta-analyses and get more reliable knowledge.

Ben Goldacre signing copies of his books
Ben Goldacre signing copies of his books.

Unfortunately this is still not standard practice, despite claims to the contrary by industry leaders (you can read more about Goldacre’s efforts to draw attention to this issue on his blog) – and there’s no legislation enforcing more transparency even though we need it urgently. As things stand, lives are lost not because we lack the necessary knowledge, but because we have failed to bring it together. The same sense of urgency doctors working on A&E experience when a patient comes in and they need to battle against time to save his or her life needs to be brought to the theoretical and abstract world of evidence. It’s very easy to lose sight of this fact, but evidence saves lives, it spares people immense suffering and grief, and it matters just as much as working directly with patients.

Lastly, Goldacre pointed out that deleting inconvenient data points is not admissible if you’re an undergraduate working on your final thesis – to do so is very obviously unethical and would get any student caught in trouble. And yet this very basic rule is forgotten once people move on to the world of medical research. Hiding negative drug trials is research misconduct, and yet companies get away with it all the time and no Royal College or professional body in medicine forbids it.

This is only a very brief summary of Goldacre’s lively and absorbing talk. He’s an absolutely brilliant speaker, every bit as brilliant as he is a writer – passionate, funny despite the grimness of his subject matter, accessible, and always engaging. In a year when I was fortunate enough to attend lots of excellent talks and author events, this just might have been my favourite. Perhaps watching his TED talk will give you an idea why:




The Geek Manifesto by Mark Henderson
Mark Henderson is the author of a book called The Geek Manifesto: Why Science Matters. Again, I haven’t read it yet, but I’m really counting on Santa – how can I possibly resist a subtitle like that? Henderson’s book is about the relationship between science and public life, specifically politics. He believes that the fact that the vast majority of politicians lack a solid grasp of science has serious consequences, and his book explains why.

Henderson began by saying that science is best understood as a way of looking at the world and acquiring knowledge – in short, as a method for making sense of the universe. Unfortunately, this aspect of science hasn’t been well understood in public life. The scientific method has immense potential, and yet it’s underexploited when it comes to policy. It’s applicable to all areas of public life, and yet we don’t use it to find out if specific strategies in, say, education or justice actually work. These are areas that could use more science, and it’s easy to do randomised and controlled trials to find out what works. But this doesn’t happen because people don’t realise how much science has to offer.

To Henderson, it’s actually unhelpful to think there’s a dichotomy between science and the humanities – science is for everyone. He clarified that when he advocates for a greater role for science in politics, he’s not calling for a technocracy where only people with scientific backgrounds are given positions of power. People’s backgrounds matter far less than their understanding of how science works. Knowing the method, more than knowing the details of what has been discovered, is what makes people able to engage with science meaningfully, and we can make a huge difference by simply embedding the process of science in how we teach it from an early age.

This, by the way, is absolutely my experience. As you probably know, I’m a humanities girl, but I feel completely comfortable engaging with science because when I was a psychology undergraduate I took an excellent research methodology class with the best professor ever. Taking that class was life-changing in a way I didn’t fully appreciate for years. It was partially thanks to my professor’s dedication to teaching us not only how to conduct research, but why the methodology mattered and how science actually works that I became the person that I am today, and also that I developed my current interests and the confidence to pursue them.

Anyway, Henderson went on to say that mainstream politician’s poor understanding of science leads to a lot of confusion. Obviously ethical considerations matter too when it comes to politics and public life, but we need to be careful not to get them mixed up with science. Opinions reached through values are often portrayed as opinions reached through scientific evidence – this is currently visible in abortion debates in both the US and the UK – which doesn’t help anyone. Again, this is not a matter of having a background in science (his own is in history), but of understanding how science works and how we can strive for objectivity despite our personal beliefs.

Returning to the topic of evidence-based practice, Henderson pointed out that measuring the impact of a particular policy is often not in a politician’s best interest. If results showed it was ineffective and they discarded it, they’d be perceived as weak and would be stigmatised, because the culture of politics runs against this. We think less of politicians who change their minds, and yet being able to change one’s mind is absolutely crucial in science. However, having humility and avoiding cheap shots when a politician changes their minds would actually make for a more sophisticated political discourse.

Towards the end of his talk Henderson also discussed the role the media have to play in all this – one of the biggest problems in how the mainstream media cover science is what he called “the fetish for phoney balance”. For example, in climate change debates on TV or on the radio, a denialist is always brought in, which gives the public the impression that there’s still a serious scientific debate around the issue. However, the media need to realise that being fair is not the same as balancing all possible views. Some views simply are factually wrong, and it’s not “fair” or “objective” to provide them a platform.

Two thousand words later, I’m done blabbing for the day. As I’m sure you could tell if you stuck with me this far, I found both events hugely enjoyable and stimulating. I can’t wait to have the chance to read and tell you all about Bad Pharma and The Geek Manifesto, which will hopefully happen early next year. (Not that I’m dropping hints for when my partner reads this post or anything. *whistles innocently*)

7 comments:

Melanie said...

So jealous! I've had the Geek Manifesto on my radar as well but not Bad Science. I've got it added to my hold list at the library now thanks! I got to see Neil DeGrasse Tyson a few weeks ago and he was awesome; science talks are the best!

Melwyk said...

Excellent post, as usual. I agree with your take on the importance of the scientific mindset, and how it is overlooked in so many ways.

Also, your point about providing a platform for the objectively wrong is such an important one! That drives me mad as well.

Ana @ things mean a lot said...

Melanie: Aren't they? There's a science festival here in the Spring and I'm super excited about it :D And definitely read Bad Science; it's an amazing book!

Melwyk: It's such a huge misunderstanding of what fairness and balance really mean, and it creates so many problems. Freedom of expression is not the same as awarding every view the same amount of credibility, no matter what the facts are.

Vasilly said...

I'm taking physical anthropology with a professor who is very passionate about science and why it's a subject for everyone (and I agree
). This post reminds me of his lectures. I'm going to see if my library has both books.

Debi said...

So wish I was there with you!!! Of course, your post is absolutely the next best thing. :) Sounds like two unbelievably incredible events.

“the fetish for phoney balance”<--Maybe my favorite part of your whole post--this whole thing has literally driven me to tears at times. Sometimes even journalists that I generally respect, like Diane Rehm, fall into this ridiculous trap. :(

Quirky BookandFilmBuff said...

Great post, Ana! I am adding all three of these books to my Goodreads wish list. The fact that unsafe drug trials are performed on lower-income people in the U.S. and citizens of developing nations makes me physically sick.

Anonymous said...

This on the surface looks fine, but by having policy dictate what is studied would be a hindrance to progress and block avenues of research not part of an existing policy. You have to have science dictate what is researched scientifically. Goldacre is wrong.

Post a Comment

Thank you so much for commenting - I do my best to reply to comments, but life conspires against it more and more often these days. But even when I don't get around to replying, every comment is read and very much appreciated.

To non-blogger users: I've been told that OpenID comments have been giving people errors most of the time, for which I really, really apologise. But you could always use the name/URL option instead, which seems to work just fine.

Thank you for reading!