Controlling for things is bad practice in general because most variables are not good controls. (https://en.wikipedia.org/wiki/Collider_(statistics) https://en.

Controlling for things is bad practice in general because most variables are not good controls. (https://en.wikipedia.org/wiki/Collider_(statistics) https://en.wikipedia.org/wiki/Collider_(statistics)) You should instead design the study so it doesn't need massaging afterwards.