posted with love. and bacon.

Correlation doesn’t imply causation

joeschmitt:

entropyas:

“Berlin and colleagues found that children who were spanked as 1-year-olds tended to behave more aggressively at age 2, and did not perform as well as other children on a test measuring thinking skills at age 3.”

—

CNN - Spanking detrimental to children, study says

Not that I want children to be randomly assigned to a spanking group or a control group to test this further, but correlation does not imply causation……

No, correlation doesn’t imply causation. Correlation sounds good, but using it could also prove that Kleenex cause allergies. A child who uses 24 boxes of Kleenex per year at 10 is 100 times more likely to have allergies than a child who only uses one box of Kleenex per year.

And anyway, the CNN story ignores the obvious conclusion from the data - that bad babies don’t get any better even with spanking. Or that awful one-year-olds tend to grow into awful two-year-olds, who at three are also stupid. Or that only stupid, stubborn parents still spank their children, and that genetically their children are more likely to be stubborn and stupid. Of course the evidence doesn’t prove or disprove any of my hypotheses without a control group.