Scientific studies, and news coverage about them, often declare that a study "controlled" for certain variables.