Correlation is not causation

Two things moving together is a clue. It does not tell you which intervention would change the outcome.

An association can arise from a causal effect, a shared cause, reverse causation, selection, or chance. Causal reasoning asks what would happen under an intervention and which assumptions or study designs justify that inference.

For example: Ice-cream sales and drownings can both rise in warm weather. Restricting ice cream does not follow as a drowning-prevention strategy from that correlation.

Keep in mind: Randomized experiments are powerful but not always feasible or sufficient. Observational evidence can support causal claims when the design and assumptions justify them.

Think about it

For an association you recently saw, sketch a third variable that could influence both quantities.

Connected ideas

Bayes rule and belief updates · The problem of induction · Feedback loops and stability

Sources and further reading

• Stanford Encyclopedia of Philosophy: Causal Models

NoteGuild starter collection · Original educational summary; source authors are not represented as platform members.