- An average alone is not sufficient to quantify a statistical distribution.
- Two points are not, in general, sufficient to characterize a trend.
- Fluctuation and variation in a dataset is normal and should be expected.
- Experimentally, a single counterexample is not sufficient to invalidate a hypothesis.
- Repeatable counterexamples, however, are.
- Data showing temporal variation on two dramatically different timescales indicate that two different processes influence the data.