Correlation vs. causation: what’s the difference?
Correlation means two things tend to move together. Causation means one of them actually makes the other happen. Correlation is easy to spot and causation is hard to prove, because two things can rise and fall together for other reasons: a hidden third thing driving both, the cause running the other way, or plain chance. Ice cream sales and sunburns climb together every summer. Neither causes the other. The sun does.
What each word means
Two things are correlated when knowing one tells you something about the other. Taller people tend to weigh more. More hours of practice tend to go with higher scores. The link can be strong or weak, and it can run either way: as one goes up, the other goes up, or it goes down.
Causation is a bigger claim. It says that if you changed one thing, the other would change because of it.
The hidden third thing
The most common way a correlation fools people is a third factor quietly driving both. Statisticians call it a confounding variable, or a confounder.
Try it below. A beach town counts ice cream sales and sunburns every month, and they move together almost perfectly. Then show what’s behind them.
Priya finds a beach town that counts ice cream sales and sunburns every month, and they rise and fall together. So should the town stop selling ice cream?
This is the same exercise that’s in the app.
Three more ways a correlation misleads
The cause runs backwards
People who see a doctor more often tend to be sicker. The visits didn’t make them sick. Being sick sent them to the doctor.
It’s chance
Compare enough numbers and some will line up for no reason at all. With enough data, a few matches are guaranteed to be coincidence.
The wrong people got counted
Survey only the people who finished a program and it looks like the program works, because the people it didn’t work for already left.
How you actually show causation
The best tool is a fair test. Split people into groups at random, change one thing for one group, and compare. That’s how medicines are tested, and it’s what websites do when they run an A/B test. Random groups spread the hidden third things out evenly, so the difference you see comes from the change.
When you can’t run a test, you look for three things instead: the cause coming first, a believable way it could work, and the same result showing up in study after study.
Questions people ask
Does correlation imply causation?
No. A correlation can be a good hint worth testing, but on its own it can come from a hidden third factor, from the cause running the other way, or from chance.
What is a confounding variable?
A third thing that affects both of the things you’re comparing, so they move together even though neither causes the other. Hot weather is the confounder behind ice cream sales and sunburns.
Can two things be correlated and causal?
Yes. Smoking and lung cancer are correlated, and study after study, from many different directions, showed that smoking causes it. A correlation doesn’t rule causation out. It just can’t prove it alone.
Learn it by doing it
Module 10, “Correlation Isn’t Causation,” is four five-minute lessons on this, starting at the beach with Priya.
Data to AI is five-minute lessons on how AI works, starting with the data on your phone, for curious teens, parents and grandparents. No math needed. The first lesson runs right here in your browser.