while designing experiements/studies/research, in which case yeah absolutely, even if thats not completely possible. Would be interested to hear Gail's perspective on this, bc she was interested when i told her about experimental econ (social sciences need!)