Suppose that there is uncertainty regarding whether it will rain tomorrow. One thing is to merely assert "I believe it won?t rain", and another is to say "I believe it won't rain, and I am quite sure of it". The latter statement is more akin to the approach statistics adopts with regard to uncertainty. An uncertain answer is not the same as an approximate answer. Statistics, from the outset recognizes uncertainty and addresses and quantifies it first-hand.

In a context where uncertainty is present, it is by definition impossible to give a correct answer with complete assurance. This is a distinguishing feature of the type of problems that statistics is able to take on. It must not be viewed as a defect of statistics. On the contrary, it is this feature which makes statistics a very useful discipline.