★ Everest Regression
November 2025
The Everest Regression Fallacy is a logical fallacy in statistical modelling that comes from over-controlling for variables.
By controlling for the variable that defines a subject, you strip the subject of its essential properties. This leads to technically correct but practically useless or even misleading results. The example that gave this fallacy its name is: “Controlling for altitude, Mount Everest is not particularly cold”. By controlling for height, you remove the essential reason why Mount Everest is extreme in the first place.
The Everest Regression Fallacy is worth keeping in mind when reading scientific papers or news articles that discuss them.
Further explanation of this fallacy can be found here.