Correlation Is Not Causation: Trend Is Not Even Close to Mechanism

“Heart disease correlates with X.” Fill in the blank. Do you believe that correlation IS causation? Careful. That’s a logical fallacy.

It crops up everywhere. “The economy does Z.” Fill in the blank. Now, that’s a really big and complex system you’re talking about. There will be a lot of correlations. Which proposed items are and aren’t legitimate causes? Current world leaders? Former world leaders? Old policy? New policy? Anticipated future policy? What if the real mechanisms which drive economy haven’t even been studied yet? What if we haven’t yet verified a mechanism?

We jump. We hurtle ourselves toward a causal conclusion. But look before you leap.

In many statistical studies, there are defined trends. So what? Without a proposed, testable, known mechanism, we have to be very careful. It is just as easy for what we think is an effect to be a cause.

In human health, correlation to causation and trend without mechanism are thin ice. Did you know that 100% of people who sleep in beds end up dying? I guess you better not sleep at all, just to be safe.
​
Correlation is not causation. Trend is not even close to mechanism.