Why Doesn't a Correlation Between Two Things Mean One Causes the Other?

The phrase "correlated with" shows up in science reporting far more often than "causes," and the difference matters. Correlation means two things move together: when one goes up, the other tends to rise or fall alongside it. It doesn't mean the first thing is actually causing the second. A third, unmeasured factor could be driving both, or the relationship could even run in the opposite direction from what seems obvious.

Observational studies, the kind that follow a group of people over time without controlling their lives, can't rule out those alternative explanations on their own. To establish causation with confidence, researchers usually need a controlled experiment: deliberately changing one factor while holding everything else constant, then watching what happens.

That doesn't make observational studies worthless. They're often the only practical way to study something over years in real human lives, and they raise questions a controlled experiment can then go test. But they stay a first step that surfaces a pattern worth explaining, not a finished, ready-to-apply answer.