Bad science & disjointed rambling
I just finished reading Ben Goldacre's Bad Science. If you've been reading his blog for a while, then there won't be much new here -- in fact, the book is almost certainly an edited collection of articles from his blog (or from the Guardian). If you haven't been reading his blog, and don't care to, then here's a summary of what you might learn, as you read through the book:
So, it's all a bit depressing. It also fits in nicely with the previous book I read, Fooled by Randomness, in which Nassim Nicholas Taleb writes about randomness in the markets. Taleb's core point is that many top traders think they know what they're doing, but they're merely lucky. At some point, they become unlucky - by this point, they have towering self-confidence and so they lose vast amounts of money. According to Taleb, this keeps happening. [Taleb, allegedly, made a fortune out of these people during the recent finance crisis]
A few years ago, I had the general world view that most stuff was sorted out. People and organisations tend to get better at things, so for anything that people have been doing a long time (e.g. trading), we should be pretty good at it. Now it turns out that many traders are actually quite bad at it, despite having lots of money to play with. And while the scientists on the front line have a good understanding of science, many of those involved in assessing the results of their work do not. So we get medical funding decisions based on politics and hysteria, rather than a rational evaluation of the evidence.
Back to _Bad Science_, Goldacre posits that it's a part consequence of the advance of technology. 50 years ago, "technology" was cars, radios, telephones, TVs. 10 year olds could build radios at home, many car owners could diagnose and fix a lot of problems, and motivated teenagers could learn all kinds of things about how telephones worked. Today, a radio is a microchip a few millimetres across, my cell phone is literally a black box, and if you take your car to a mechanic the first thing they'll do is plug it into their computer to interrogate the car's computer. Understanding modern technology is literally beyond almost everyone's capabilities, so we're really well into the territory of Clarke's third law.
So the complexity of modern life makes us reliant on the people who do know how things work. Hence people get into the habit of listening to experts - and if there's money in it, someone can always find a self-styled "expert" to push a view not supported by evidence.
(a recent example in NZ is the guy who took intravenous vitamin C and got better. The latter event followed the former in time, and we are encouraged to think there is causation here. But will the newsmedia cover the story of the people who took intravenous vitamin C and died?)
- Homeopathy, acupuncture, ear candling, etc. are all placebos. This probably isn't news to anyone; the main point of the chapter is to illustrate the scientific method by showing how we know they are only placebo.
- Saying "only placebo" is not the same as "does nothing"; the placebo effect is weird and powerful.
- All moisturisers are the same.
- Big pharma is evil, but maybe not for the reasons you think. Pharmaceutical companies make the biggest profits on drugs that are still under patent. Unfortunately, medical science has gathered all the low-hanging fruit, and so coming up with new drugs that are better than existing drugs is very hard. Pharma will run slightly-dodgy trials, will suppress trials that don't give the right result, and rely on marketing rather than hard evidence to convince doctors and consumers. Oh, and there could actually be some easy results still available - but only in the third world. Poor countries can't afford expensive drugs, so their medical problems get ignored.
- Nutritionalists are just another group of alternative medicine practitioners. Good nutrition is really straight-forward: eat a varied diet including fresh fruit and vegetables and dietary fibre, don't overeat, don't drink to excess, don't smoke. Done. Everyone knows that, so nutritionists, to justify their existence, need to make things seem more complicated than they really are. [incidentally, anyone can call themselves a nutritionist - there's no degree requirement] Every time a "health supplement" has been tested in a controlled trial, it's turned up no better than placebo (well, there was one case when over 10,000 Finns at high risk of cancer were given either antioxidant supplements or placebo. The trial was stopped on ethical grounds because the group on the antioxidants were suffering 45% more likely to die of cancer than the control group..).
- Finally, whom do we blame for all the quackery? Goldacre piles most of the blame at the door of the news media. The media want exciting stories, which means they want (a) miracle cures, and (b) hidden death risks. Partly, they want this because those stories used to be genuine. Cigarettes and asbestos were definitively revealed as killing millions. Vaccines knocked out smallpox and polio. Etc. Unfortunately, these days breakthroughs are more likely to be "this drug is the same as that one, but the side-effects are not so bad", or "this drug is 5% better than that one". And the risk factors have turned out to be more complicated -- or more boring. I expect I can list the major health risk factors today: poor diet, lack of exercise, over-drinking, cold/damp housing. No magic drug will solve those problems.
Beyond this, the news media thinks their readers don't understand science. So when they do print science subjects [aside: ever noticed how science stories in the Dom Post often turn up in the "odd stuff" section?], they don't print the evidence. Rather, they portray it as a pronouncement from an authority figure. Which means if the story conflicts with anyone's world view (e.g. stories on alternative medicine), they can just bring in a different authority figure to issue a conflicting statement. Presented with two authorities and no evidence, what are the readers to think? They pick the view that best accords with their world view, because they have no reason to choose otherwise.
So, it's all a bit depressing. It also fits in nicely with the previous book I read, Fooled by Randomness, in which Nassim Nicholas Taleb writes about randomness in the markets. Taleb's core point is that many top traders think they know what they're doing, but they're merely lucky. At some point, they become unlucky - by this point, they have towering self-confidence and so they lose vast amounts of money. According to Taleb, this keeps happening. [Taleb, allegedly, made a fortune out of these people during the recent finance crisis]
A few years ago, I had the general world view that most stuff was sorted out. People and organisations tend to get better at things, so for anything that people have been doing a long time (e.g. trading), we should be pretty good at it. Now it turns out that many traders are actually quite bad at it, despite having lots of money to play with. And while the scientists on the front line have a good understanding of science, many of those involved in assessing the results of their work do not. So we get medical funding decisions based on politics and hysteria, rather than a rational evaluation of the evidence.
Back to _Bad Science_, Goldacre posits that it's a part consequence of the advance of technology. 50 years ago, "technology" was cars, radios, telephones, TVs. 10 year olds could build radios at home, many car owners could diagnose and fix a lot of problems, and motivated teenagers could learn all kinds of things about how telephones worked. Today, a radio is a microchip a few millimetres across, my cell phone is literally a black box, and if you take your car to a mechanic the first thing they'll do is plug it into their computer to interrogate the car's computer. Understanding modern technology is literally beyond almost everyone's capabilities, so we're really well into the territory of Clarke's third law.
So the complexity of modern life makes us reliant on the people who do know how things work. Hence people get into the habit of listening to experts - and if there's money in it, someone can always find a self-styled "expert" to push a view not supported by evidence.
(a recent example in NZ is the guy who took intravenous vitamin C and got better. The latter event followed the former in time, and we are encouraged to think there is causation here. But will the newsmedia cover the story of the people who took intravenous vitamin C and died?)