Skip to content
Probability & Statistics

Statistical Thinking

Correlation ≠ Causation

7 min+70 XPStep 1/10

Two curves that move together

Ice cream sales and drowning deaths rise and fall together — tightly correlated. Does ice cream cause drowning? No: summer causes both.

Correlation measures how two variables move together. Causation says one produces the other. Correlated variables can be:

  • Genuine cause → effect (smoking → cancer)
  • Effect → cause, backwards (sick people take more medicine)
  • Both driven by a hidden third factor — a confounder (summer → both)
  • Pure coincidence (margarine consumption vs Maine's divorce rate — a real, spurious pair)

analogy

Correlation is two people dancing in sync. Causation means one is leading. Often there's a DJ — a hidden confounder — making both move.