Sunday, 7 December 2014

Et tu, Randall Munroe?

You all know who Randall Munroe is, right? Or at least if I put the letters X, K, C, and D together in that order you know what I’m talking about? Good.
xkcd is one of the science-y-est things on the internet. But even Munroe can slip on the old “I bet those scientists didn’t anticipate the potential flaw in their method that I spotted when I read a lay account that didn’t go into all the fiddly details about controls and things” banana-skin once in a while. Here’s the strip I’m talking about:
Our fMRI study found that subjects performing simple memory tasks showed activity in the parts of the brain associated with loud noises, claustrophobia, and the removal of jewellery.
Munroe’s mouse-over caption reads “They also showed activation in the parts of the brain associated with exposure to dubious study methodology, concern about unremoved piercings, and exasperation with fMRI techs who won’t stop talking about Warped Tour.”

Um, Randall? Do you know what the f in fMRI stands for? It’s “functional”. Functional magnetic resonance imaging differs from boring old magnetic resonance imaging in the software it uses, and what that software basically does is record changes in brain activity (well, in blood flow to particular parts of the brain) between the moment before the experimental stimulus and the moment after it. Any activity attendant on the experience of being in an fMRI machine, per se, will not change. It will be the same on both sides of the stimulus and so will cancel out.
Which doesn’t of course mean that science can’t or shouldn’t be criticized by non-experts. Just check whether they’ve figured out what you’ve figured out before you start telling people that they haven’t, OK?

No comments:

Post a Comment