Wednesday, November 25, 2009

Studies show you ought to be dead

I did enjoy this. The Listener's Nutrition column featured an interview with the PM's science advisor, Peter Gluckman. Here's an exerpt:

“Science is not simple,” he says, “and particularly the issues that tend to impact on the public; they’re usually involving complex science. Things like human biology and nutrition, for example, are complex issues. Nutrition represents very complex biology, which is projected by people as very simplistic biology. It’s complex because you have food, which is more than ingredients. Then you have the complexity of the biological organism that eats it, namely the individual, and every person has a different genetic, developmental and physiological make-up. Therefore the interaction between that individual and what they eat can be quite variable.”

Which is why bollocks like the Listener's nutrition column, or Petra Bagust's finger-wagging at us on TV, in which we're told to eat this or not eat that because "studies show..." is exactly that - bollocks. Why? Well, here's the Listener on Gluckman again:

“The media and others are bad at communicating concepts of probability and risk,” says Gluckman, who dislikes the use of phrases such as: “eating food X doubles your risk of disease Y”.

“You’ve got to look at what that means. It might mean nothing if it’s shifting it from one case in 200 million to one case in 100 million. And yet we regularly see that kind of reporting, particularly in relation to food.”

Using a numerator and denominator, rather than percentages or doubling terms illustrates the true impact. “If something increases your risk from one in a 100 to one in 10, [that] is a very big increase in risk. But an increase in risk from one in a million to one in 100,000, which is equally a tenfold increase in risk, is still a very rare risk.”

And yet, how often do you see supposed "scientists" telling you not to eat this, that or the other on the basis of exactly those kind of very rare risks? How often? Do you know? I do - the answer is "Fucking."

I read Gluckman's comments as pretty much an admonishment of the interviewer by Gluckman, but she doesn't seem to have shared that view, given that she published them. All I can say is, keep Gluckman's words firmly in mind during the next lot of squawking from Professor Rod Jackson (the man who gave us "there should be a health tax on butter" and "[vitamin supplements] should be prescription-only drugs.")


Anonymous said...

He should keep his own words in mind, after all didn't he recently say that he had spent 3 weeks studying global warming and it was a scientific fact

Anonymous said...

ALL Gluckman's comments are designed to demostrate that the audience are not as smart as the Gluck! Try attending some of his seminars... ;(

He does have a point about poor media reporting of stats though. But bthat is because most media are arts graduates, and haven't done stats since senior high school (and they mostly didn't follow that!).