I'm not sure what you are getting at with your school example.

It's quite simple. Clearly there are times to consider the individual and clearly there are times when we should consider larger groups.

If Town A had a cancer rate 5 times higher than the national average surely you wouldn't say, "Well, look, those people are individuals, and an individual's cancer risk is affected by lots of things -- diet, exercise, genetic predisposition... to heck with looking for differences between Town A and Town B."

Clearly you'd want to start investigating things that affect Town A on a town-wide basis.