If the sample results seem inconsistent with what was expected, then the assumption about the population is probably true.

Hypotheses are made about the population

In the sample, we know exactly what to expect

The variation in statistics from sample to sample is called sampling variation