What makes a sampling method biased?

1)The sample is really big

2) the sample value is always the exact same as the population value

3)The sample value is consistently off in the same ddirection compared to the population value

4) The sample value might be too high or two low compared to the population value. It will be different for each sample