Commentary: On representativeness.

and more dependent on locally varying modifying influences, the ultimate aim of biological research on humans or other species, is like that of physics, to be able to make general statements about nature. Paradoxical though it may seem, statistical representativeness leads to particular statements about the world, not general statements about nature. As initial steps, surveys may help to seed hypotheses and give a push toward scientific understanding, but the main road to general statements on nature is through studies that control skillfully for confounding variables and thereby advance our understanding of causal mechanisms. Representative sampling does not take us down that road.