You're conflating again. He/she made a valid point. Legit studies have to control for factors and input.

Correlation is not causation.