A A A

Can we justify non-probability sampling methods when there are known characteristics of the population?