You can't demonstrate causation by controlling for things, because of collider bias. https://en.wikipedia.org/wiki/Bad_control https://en.wikipedia.org/wiki/Ba

You can't demonstrate causation by controlling for things, because of collider bias. https://en.wikipedia.org/wiki/Bad_control https://en.wikipedia.org/wiki/Bad_control You need better methods like an intervention study or a natural experiment to show causation.