OK, this one wins the worst-figure contest:
The authors of this paper (Cournot, M. et al.Relation between body mass index and cognitive function in healthy middle-aged men and women. Neurology 2006, 67, 1208–1214. or read the PDF: cournot_fat_stupid_neuro_2006.pdf) try to argue that there is a negative correlation between body-mass index (BMI) and cognative function (i.e. fat people are stupid). They report the correlations (such as -0.008 +/- 0.13) that are the equivalent of saying, “There are 3 people in this room … plus or minus 300.”
I swear to god, this is a quote from the paper describing that figure (this is not taken out of context, read it for yourself):
This cross-sectional association between BMI at baseline and cognitive scores persisted after adjustment for age, sex, educational level, physical activity, and region of residence and had a linear shape (figure).
Linear shape?!? Looks more like a big fat (pun intended) blob to me. Isn’t that awesome? The funniest part is that they worked so hard to adjust for all these different parameters, and still got a complete scatter. The authors must be fat, ’cause they’re pretty stupid. I can’t believe this stuff can get published! This is why people are starting to claim that most published research findings are false.
Too bad I already made the mugs for the Chemical Physics Journal Club this year, because that plot would have been perfect!
[I have to thank Spectroscope for bringing my attention to this travesty. I also have to thank Retrospectacle for reminding me why the softer sciences have a bad reputation; but my housemate thinks Shelley's hot, so I won't be too mean.]