n = 12 ?

So I see other people are writing a bunch of stuff about Andrew Wakefield, the guy who published a study linking autism and MMR-vaccination. Here’s an article in the British Medical Journal about the case. I haven’t read much about the stuff before, maybe I should have.

It turns out Wakefield’s study was a study based on 12 individuals. 12. I did not know that.

Ok, so this is why there’s no way I’m giving him the larger part of the blame for what has happened here. If you make a decision as risky as one exposing your child to multiple dangerous and easily preventable diseases, you’d damn better have a better reason than a study on 12 individuals. If not, you’re a gullible moron.

Wakefield is a fraud, but even if that’s the new thing others add to this story today, it’s not what I’m focusing on. How the hell did a study on 12 individuals ever get to be considered important enough to have any kind of impact on, well, anything? Who were those fools who were giving a study like this any kind of credence? Well, there were a lot of them (wikipedia):

“The claims in the Lancet article were widely reported;[7] vaccination rates in the UK and Ireland dropped sharply,[8] which in turn led to greatly increased incidence of measles and mumps, resulting in a few deaths and some severe and permanent injuries.[9]

Between 1997-1998 and 20001-2002 the MMR vaccine coverage of 2-year-olds in the UK fell from appr. 91 percent to 84 percent. The number of measles cases in the UK was more than 10 times as high in 2007 as it was in 2000 (children who’re not vaccinated don’t necessarily get sick, if they get sick, right away).

I’m reminded of the recent case of a type 1 diabetic child who died because the parents were too moronic to take him to the doctor, relying instead on the power of prayer to heal the kid. It wasn’t the first time that particular story had played out, nor will it be the last. If you’re that stupid, don’t have kids.


January 6, 2011 - Posted by | Diabetes, Medicine, Studies, stupidity


  1. I’ll see your sample size of 12, and raise (diminish?) you with a sample size of 2(two): . I blogged about it a while ago, and here are my thoughts:

    “Stop the presses, newsflash, groundbreaking research unveiled: people are afraid of losing money, and it has something to do with their brains.

    The scientific feat has been achieved by studying 2 (yes, that is indeed two) subjects – both female, both with a serious genetic defect – and comparing their behavior in a lab test with that of 12 (yes, that’s twelve) other people.

    You know, it does not matter that economists have suspected something like that since Harry Markowitz proposed the portfolio selection theory in 1952. It matters even less that statisticians have developed this weird concept of representative samples in order to deal with inference from anecdotal evidence. What does matter is that you can get a nice grant and career advancement opportunities by documenting quasi-research that shows support of a well-known fact on OPM (other people’s money). Considerations such as, say, the subjects’ incomes and wealth (actually, full financials – balance sheet, income statement, and cash flow ) – known to affect risk aversion – are just trifling, nitpicking details. Why should we correct for exogenous variables, if the “research” will create buzz anyway? Just read the damn paper, and clap, rubes. Oh, yeah, and write a check for a follow-up study – who knows, we may discover something even crazier – say, that people (probably because of their brains, but you never know) enjoy winning money…”

    Comment by Plamus | January 9, 2011 | Reply

  2. To be fair, economists haven’t talked much about which effect the amygdala might have on risk aversion. But you’re right, I feel the media would do most of their readers a favour if they simply stayed far, far away from that sort of small-n studies in general – there are plenty of much larger studies with much greater statistical power that go un(/-under)reported all the time, and the sensationalism associated with small studies like these often causes readers to downright misunderstand which conclusions should be drawn from them, if any, even if the researchers are very careful with their choice of words when they’re communicating with the press (which they most often aren’t).

    However, of course the medias need the ‘buzz’ because buzz is what sells papers/adds. None of the traditional media seem to me to have a strong incentive to ‘do a proper job’ reporting on science, quite the contrary; they have an incentive to tell people about small (and often flawed) studies with Big Conclusions, studies about Individuals, not studies about Data. Studies where the professor isn’t too busy researching or lecturing to tell yet another journalist what all those equations were about and what was really going on in that paper. Studies that do not take long to digest but still might spark peoples’ interest. Stuff with a knowledge component low enough for the journalist not to be forced to use the first two pages to write about stuff you need to know in order to understand what the study is about.

    I’ve sometimes commented on the lack of knowledge about statistics most journalists have. But all that kind of stuff doesn’t really matter, does it? It’s about the money. It’s about what’s easy for the journalist and it’s about what people want to read. All that other stuff…

    People read newspapers primarily because they want to be entertained, not because they want to be informed. Very few people are willing to pay a lot to obtain accurate reporting on science. Especially since quite a few brilliant science-bloggers and researchers deliver the goods for free online. I have to keep remembering this every time I feel like yelling at the journalists – they just do what they gotta’ do to stay in business.

    Comment by US | January 9, 2011 | Reply

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: