If a sample "fails to cover the entire population", it's biased. In my opinion "Sample is not random" and "Sample is biased" should be the same point.