Ben Goldacre on MRI for market research

"With science, you publish a description of your experiments, in an academic journal, so that people can see what you've done, not your interpretation. That is why academic journals exist, instead of just newspapers..."

- Ben Goldacre, writing in the Guardian

This week, Britain’s favorite gadfly tackles the science behind brain imaging for market research purposes. He was responding to an Op-Ed piece written in The New York Times by the people behind the company FKF Applied Research. FKF’s piece described a study claiming to elucidate what the American electorate likely feels about the latest crop of Presidential hopefuls.

For those of you who haven’t heard of FKF, its website describes itself as ‘a high-tech, next generation business intelligence firm selling functional Magnetic Resonance Imaging (fMRI) brain scan-based research to Fortune 500 clients. By looking inside consumers' minds with the fMRI, FKF provides valuable and never-before-available insights into almost every aspect of its clients' businesses with a specific focus on brand marketing and advertising.’ Its methods, we are told, are scientifically sound and empirically precise.

The study described in the NYT purports to get inside the head of voters with uncanny detail; for example, we are told that scans suggest “Mrs. Clinton may be able to gather support from some swing voters who oppose her if she manages to soften their negative responses to her. But she may be vulnerable to attacks that seek to reinforce those negative associations.”

If it seems a bit far-fetched that emotions of this subtlety could be gleaned from something as crude as an MRI, you’re not alone. Goldacre describes how current MRI evaluations rely on very simple questions that eliminate as many variables as possible, and that interpretation very much depends on what sort of question is asked. It is therefore difficult to evaluate tests that are likely to stimulate more than one kind of emotional response or that could stimulate different emotions whose centers happen to be closely associated in the brain. Given the complexity, this sort of study cannot be taken at face value without the readers being given access to their methodology.

You can read the rest of Goldacre’s article here.