EN BG

Correlation vs Causation

Two things happened at the same time. That's it. That's all you know.

Correlation means two things move together — when one goes up, the other tends to go up too. Causation means one actually produces the other. The gap between those two is where most bad reasoning lives. Ice cream sales and drowning deaths both rise in summer. Nobody thinks ice cream causes drowning. The missing variable is obvious: hot weather. But when the missing variable isn't obvious — when it's hidden behind complex statistics or an impressive-sounding study — you accept the link without thinking twice.

Media thrives on this confusion because "X linked to Y" gets clicks while "X and Y both happen to increase under conditions we don't fully understand" does not. Drug studies exploit it too. A pharmaceutical company can show that people who take their supplement also report better health — without proving the supplement did anything. Maybe healthier people are more likely to buy supplements in the first place. The correlation is real. The causation is invented.

The honest move is unglamorous: it requires controlled experiments, isolated variables, and the patience to say "we don't know yet" when you don't. Every time someone presents a correlation as proof, ask the boring question — what else could explain this? The answer is almost always "a lot of things." That is not a failure of knowledge. That is where real knowledge begins.

References