One way to do this is demonstrate that a suitable manipulation and/or measure allows us to influence and/or predict some outcome that people care about. Another is to show that particular factors or influences are more powerful and important (or in some cases less so)—either in absolute terms or relative to other factors—than recognized by lay social observers or even supposed experts.5