Sampling Bias

Sampling bias is a type of non-sampling error that arises from a flaw in the sampling plan itself. It occurs when the plan is designed in such a way that some members of the target population could not possibly be included in the sample — that is, certain units have no chance at all of being selected.

Because a whole segment of the population is systematically shut out, the resulting sample cannot be representative of the population it is meant to describe, and the estimates drawn from it will be distorted no matter how carefully the rest of the survey is carried out. This is what makes non-sampling errors like sampling bias more troublesome than ordinary sampling error: whereas sampling error can be reduced simply by taking a larger sample, a larger sample does not cure sampling bias — if a group is excluded by the very design of the plan, enlarging the sample only draws more units from the same restricted pool a …