Maybe Results Imaginary

Scott Hardie | November 29, 2014
Kelly and I were just talking about this the other day, and then Friday's XKCD hit on the same idea: Why does one part of the brain lighting up in an MRI "prove" anything about what the subject is thinking?

The method is so commonly used that we now claim to know what dogs are thinking just because there's apparent brain activity when sounds are played. Does that prove what the researchers think it does about dogs processing the sound and interpreting emotion within it? The dogs' brain activity seems that way, but we have no way of knowing what the dog is thinking; it could be thinking about sniffing another dog's butt at that moment in time for all we know. Correlation is not proof; it's just correlation.

This trend in science news towards reporting what the brain is thinking just because an MRI lights up can subtly undermine critical thinking. I think it's another sign of us placing faith in technology because it can tell us what we want to hear. It's like polygraph machines, aka lie detectors: They measure heart rate and blood pressure and so on, but do those physical symptoms actually correspond to lying? Studies have shown that they do not, but we shouldn't need studies to tell us that we're making a logical leap when we start imagining connections that we wish existed.

What do you think?

Samir Mehta | November 29, 2014
[hidden by request]

Scott Hardie | December 13, 2020
I'm glad to hear that science is turning away from brain scans, even if it's for a bunch of other reasons that sidestep the fundamental problem that I mentioned above.