MegaMaester

Critical Thinking · Lesson 6

Correlation vs Causation

beginner15 min · 9 cards
The big idea
🔗

Together is not because

Two things moving in step tells you something is going on — not that one is causing the other.

The hidden third factor

What a confounder does

Hot weatherIce creamDrownings?
Hot weather drives both ice cream and swimming. Neither touches the other.
Rule these out first

Three innocent explanations

Tap each.

Try it

Confounder Finder

Ice-cream sales rise when drownings rise.

What best explains the pattern without one causing the other?

A correlation is a reason to investigate, never a conclusion.

Diagnose why each pattern is not evidence of cause.
Key term

What would settle it?

How do you actually show cause?
tap to reveal ↻
Intervention
Change X deliberately and watch Y — a randomized controlled trial. Where that is impossible, cause is built from mechanism, dose-response, and ruled-out confounders across many studies.
Nuance

Reading headlines

Think like a Maester

If the study observed people rather than intervening, the honest phrasing is “is associated with”. Anything stronger was added later.

Recap

Key takeaways

  • Correlation describes a pattern; causation claims a mechanism.
  • Confounding, reverse causation, and coincidence each produce correlation without cause.
  • Spurious correlations are inevitable when enough data is compared.
  • Establishing cause needs temporal order, mechanism, ruled-out confounders, dose–response, or intervention.
  • Observational studies support "associated with," not "causes."
Quick check

What best describes critical thinking?