Confounding variables

Suppose you're designing a clinical trial to determine whether people from Inner Mongolia are better or worse swimmers than those from Outer Mongolia.