“Breaking: New study finds drinking expensive alcohol linked to longer lives and drinking cheap alcohol linked to shorter lives,” the author Keith Payne tweeted recently.
“Scientists desperately trying to identify the features of alcohol molecules that could explain this puzzle,” he added.
The joke, of course, is that people who drink expensive alcohol very likely also enjoy access to better healthcare and can afford more nutritious food, among other things.
It is a good comedic illustration of why correlation is not causation.
But it is all too easy to fall into this trap in business.
In marketing we love to keep an eye on the numbers. When launching a new campaign for an outdoor pizza oven, for example, there is nothing more satisfying than watching sales increase as brand and product awareness grows.
And often there is a direct link between the marketing and the sales.
But sometimes marketers have to admit there might be other factors at play.
Maybe the new campaign coincided with unusually warm weather. Or perhaps a popular new Netflix show features a character that loves nothing more making pizza in an outdoor oven.
Maybe both these factors caused the spike in pizza oven sales, and not the ad campaign. Until you really dig into the detail, it is difficult to know.
Another example is an e-commerce business that sends a 25% off coupon code to its email newsletter subscribers.
After looking at the results it might be tempted to attribute any sales that used the coupon as being a sign the campaign worked. Maybe the business should send out more coupon codes and drive more sales.
But before doing so it is worth checking if the coupon users were regular customers that were likely to have made a purchase anyway.
Rather than increasing sales, the retailers might have just cut into their profit margins for no good cause.
And worst of all, it may have inadvertently taught its best customers to hold off on making a purchase until they have a coupon.
“It is said that there is a correlation between the number of storks’ nests found on Danish houses and the number of children born in those houses. Could the old story about babies being delivered by storks really be true?” says economist Tim Harford.
“No. Correlation is not causation. Storks do not deliver children but larger houses have more room both for children and for storks.”