Wednesday, September 08, 2004

Correlation is not Causation

It's an extremely common mistake for laypeople to make, but scientists should know better:
Keeping a diary is bad for your health, say UK psychologists. They found that regular diarists were more likely than non-diarists to suffer from headaches, sleeplessness, digestive problems and social awkwardness.

Now surely this is pointing out the obvious: but maybe their suffering is what caused these people to start writing, rather than the other way around?

She claims that her study is the first to investigate subjects who write of their own free will. In most other studies, volunteers are actually asked to write about traumatic experiences in a systematic way.

Yeah, there's a reason for that: controlling the variables. If you take your sample population, and randomly assign them to experimental vs control groups, then the only consistent difference between the two groups will be that variable which the experiment is testing. So any difference in results can be put down to just that variable. However, if you do a broad correlational study, rather than a proper experiment, there are all sorts of other variables that could interfere with the results. In this case, those who are most traumatised might feel the most need to write about their experiences.

Statistically, the diarists scored much worse on health measures than the non-diarists. And worst affected of all were those who had written about trauma. “They were most susceptible to headaches and the like,” says Duncan.

Let me get this straight: We have a bunch of diarists who are also trauma victims. They were found to be more likely to have headaches (etc) than your average student. And the psychologists are blaming this on the writing?

Perhaps New Scientist has misrepresented the study, but as it stands, these psychologists sound downright stupid. At least they show a glimmer of sense towards the end of the article:
But she acknowledges that her experiment could not demonstrate which came first - the diary writing or the health problems.

In a forthcoming experiment, she hopes to explore this by asking volunteers new to diary writing to report exclusively positive or negative things, to see if the health of the two groups diverges.

And here I thought she was going to go on to prove that wet pavements cause rain, or ice cream causes drownings.


Post a Comment

Visitors: check my comments policy first.
Non-Blogger users: If the comment form isn't working for you, email me your comment and I can post it on your behalf. (If your comment is too long, first try breaking it into two parts.)

Note: only a member of this blog may post a comment.