MEPX
Chapter 8 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 8 of 10

Regression to the mean

A cause that is not there.

The pattern

Extreme measurements tend to be followed by less extreme ones with no intervention at all. Treat the worst performers and they improve; so would they have anyway.

  • Any programme targeting extremes will appear to work.
  • This is why a control group is needed even when the effect looks obvious.

Everyday examples

Sports form, illness after a bad day, and pupils who did unusually badly on one test. Praise and criticism both appear to work for exactly this reason.