Teach nonscience majors uncertainty?

Is it a complete waste of time to try to teach experimental uncertainty to non-science majors?

Getting science majors to understand the idea of uncertainty is tough enough. However, I can usually get some significant fraction of a class to get the general idea. But non-science majors?

I'm not talking about having them propagate uncertainties or otherwise calculate anything. But it seems to me that they should be able to grasp that limitations of your measurement give you some idea as to whether your percent error is big or small.