Sceptic Schism | Issue 19

Sceptic Schism | Issue 19

Correlation ≠ Causation

I magine you buy a new pair of assless rubber underpants (I’m not judging) and you wear them the day that you happen to get your dream job at the waterslide-testing and kitten-cuddling factory. On the same day, you get asked out by the guy from the Old Spice ads, and land a six-digit book deal based on your exceptionally good online arguing skills. You decide to credit your undies for your “good luck”, and wear them every day until your bum is rubbed raw and you stink of deteriorating rubber. But you have made the mistake of believing that a coincidental thing is causing something unrelated. Your mistake is frightfully common.

The history of medicine is mostly a history of clueless people doing terrible things to the human body. Many ailments clear up on their own, and will heal themselves despite, not because of, all the things we try to fix them with. Mercury was used to treat gonorrhoea for decades, although it was ineffective and harmful. Bloodletting was used to treat pretty much everything until well into the nineteenth century, despite being terrible for anybody’s general health. But it was survivable, and so it was used for centuries by well-meaning doctors who thought patients who recovered were benefitting from having the blood drained from their already weakened bodies.

If you and your community are desperate for rain and you sacrifice an animal, it is almost certain that it will rain at some point in the future after you do it. This does not mean your sacrifice caused the rain.

 There is pretty good evidence that products for the prevention and treatment of the common cold are useless, yet we continue to buy millions of dollars worth of them every year. Anti-aging creams are basically expensive moisturisers with different names. I found a delightful website of “spurious correlations”, which maps unrelated things against each other on graphs so it looks like they have a common cause. For example, “Number of people who drowned by falling into a pool” correlates with “Number of films Nicolas Cage appeared in”, “Age of Miss America” correlates with “Murders by steam, hot vapours and hot objects,” and “People who drowned after falling out of a fishing boat” correlates with “Marriage rate in Kentucky”. Coincidences? Well who knows? Maybe don’t get married in Kentucky, just in case.

This article first appeared in Issue 19, 2015.
Posted 2:18pm Sunday 9th August 2015 by Wee Doubt.