Spurious 

Author Tyler Vigen runs the website Spurious Correlations. On it you'll find numerous intriguing examples of beautifully correlated data, such as the connection between U.S. per capita margarine consumption and the divorce rate in Maine, or the number of people who drowned by falling into a pool and the number of films featuring Nicolas Cage. These examples, often termed Pure coincident, demonstrate that the first step in analysing cause and effect is acknowledging that two variables may only be coincidentally matched. The one doesn’t always cause the other.

Backward 

In an article in The Wall Street Journal, Assistant Professor Luis Martinez, from the Harris School of Public Policy, discusses the relationship between economic expansion in autocracies and the light from such countries as seen by satellites in space. He wrote that what the satellites observe provides us with a far more accurate picture of an autocracy’s gross domestic product than the information often released by its leaders. 

‘As an economy expands,’ he says, ‘there are more houses, factories, more streetlights. All of that produces light.’ For him, it’s evident that economic expansion leads to the production of light, rather than the other way around. This is Backward causation. Or to put it another way: when determining cause and effect, it’s important to get them in the right order.

Confounded

For his book, Everybody Lies, data scientist Seth Stephens-Davidowitz took a look at Google Correlate. This was a Google program – now discontinued - that brought together information about multiple search terms. He discovered that the most tightly correlated search by unemployed individuals was not for unemployment offices or new jobs, but rather for a famous pornographic website. But he didn’t think there was a direct cause and effect relationship between being unemployed and searching for pornography. He thought the explanation was much simpler: that some people were simply passing the time. 

It’s an example of an almost-invisible Confounding factor, and it’s so frequently missed that analysis results are often interpreted without taking it into account.

Relevant to creatives?

No doubt by this point, you’re wondering what any of this has to do with the creative community. It’s this. If we want to accurately evaluate the impact, and the return-on-investment, of new designs, advertising campaigns, and even in-house design capabilities, then we need to understand the distinction between correlation and causation. We need to be able to separate genuine, valid conclusions – however uncomfortable we might find them - from more eye-catching but ultimately spurious findings.

To do that, we need to master the art of precision data mining. We need to learn how to properly study numbers and spreadsheets – and to draw the correct conclusions from them. We need to learn not how to manipulate data, but how to effectively demonstrate cause and effect. Only then will we be able to make a distinction between correlation and causation.

For, as Ralph Waldo Emerson once said, ‘Shallow men believe in luck. Strong men believe in cause and effect.’