The neuroscience blogosphere and science reporting in mainstream media has been abuzz lately over a paper published by Vul, Harris, Winkielman, and Pashler1, titled "Puzzlingly High Correlations in Studies of Emotion, Personality, and Social Cognition"2. A really amazing and clear analysis of the hype/media coverage by ScienceBlogger Neuroskeptic can be found here: "Voodoo Correlations" in fMRI: Whose Voodoo? This makes the paper and the issue really easy to synthesize, and keeps a good grasp on the scientific principles for which research like social cognition and fMRI should always strive. Neuroskeptic even managed to make confusing statistics really clear! I highly recommend reading this one, particularly if you're into fMRI studies at all.
MadCog
1 You may recognize Vul and Pashler as the authors of "Measuring the Crowd Within", a paper on "the wisdom of crowds": the averaged guess of a group of people is more accurate than the guess of any one person. That's a fun interesting quick one.
2 This was called "Voodoo Correlations in fMRI Studies" in the Neuroskeptic article, I'm guessing that must have been an earlier title.
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment