In estimating from a random sample, the difference between the estimator and the parameter can be written as the sum of two components: bias and sampling error. The bias is the average error of the estimator over all possible samples. The bias is not random. Sampling error is the component of error that varies from sample to sample. The sampling error is random: it comes from "the luck of the draw" in which units happen to be in the sample. It is the chance variation of the estimator. The average of the sampling error over all possible samples (the expected value of the sampling error) is zero. The standard error of the estimator is a measure of the typical size of the sampling error.