MEPX
Chapter 3 of 10All chapters
  1. 1What correlation measures
  2. 2The five explanations
  3. 3Confounders
  4. 4Experiments
  5. 5Natural experiments
  6. 6Weighing causal evidence
  7. 7Selection effects
  8. 8Regression to the mean
  9. 9Language and reporting
  10. 10Acting under uncertainty

Chapter 3 of 10

Confounders

The third factor.

How it works

A confounder influences both variables, creating an association between them with no direct link. Summer causes both ice cream sales and drownings.

  • Adjusting statistically requires knowing about and measuring the confounder.
  • Unknown confounders are the reason observational evidence stays weaker.

Colliders

Controlling for a variable that both influence can create a false association. Adjustment is not automatically safe, which is why a causal diagram should come before the model.