Plain language, shorter sentences, no jargon.
Correlation vs Causation: The Most Common Statistical Mistake
Why two things moving together doesn't mean one causes the other
Ice cream sales and drowning rates both peak in summer. Should we ban ice cream to save lives?
Start here: make a prediction before reading
A correlation means two variables tend to move together. Causation means one variable actually causes changes in the other. These are completely different claims, and confusing them leads to disastrous policy and reasoning.
The ice cream / drowning correlation exists because both are driven by a third variable: hot summer weather. When you ignore confounding variables, every correlation looks like a potential cause.
Headline: 'Coffee linked to lower cancer risk.' Does that mean you should drink more coffee? Not without seeing experimental evidence. Training yourself to ask 'is this correlation or causation?' is one of the most powerful critical thinking habits you can build.
Quick Check
Which of these best captures what you just read?
It's okay to be wrong. The thinking process is what matters.
Quick Check
Question 1 of 3Why do ice cream sales correlate with drowning rates?