MEPX
Chapter 7 of 10All chapters

Chapter 7 of 10

Cause and effect

The most common overreach.

Alternatives to consider

Before accepting that A caused B, consider chance, reverse causation, a third factor driving both, and selection deciding who was in the data at all.

  • Only randomisation reliably rules out the third factor.
  • Before and after is not evidence: things change for many reasons.

Regression to the mean

Extreme results are usually followed by less extreme ones with no intervention at all. Much of the apparent effect of treating the worst cases is exactly this.