Thought of the week, from Randy Pherson: People need to be "N-sensitive"—to pay attention to how much evidence they have (the value of "N") before they draw conclusions. If you have one friend who eats a pickle and then comes down with the 'flu, that doesn't mean that pickles cause 'flu. If millions of people smoke cigarettes and thousands of them get cancer, however, that can be extraordinarily strong evidence of causation (or at least correlation). It all depends on getting a big enough N for the statistics to be significant.
A better word for it: N-aware!