List Sampling

List sampling is one of the basic ways that survey samples can be created. The basic concept of list sampling is deceptively simple. The process is to choose a subset of the elements (the sample) from a listing of all elements (the sampling frame) using a specific selection process. The selection process may have several features, for example, sampling with replacement or sampling without replacement.

In list sampling, as in other sample selection processes, issues arise about whether the sample estimate is an unbiased and reliable estimate for the characteristic or attribute in the full list of elements. Bias and reliability are measures of how well the estimator for the attribute computed using list sample data corresponds to the true value for the attribute in the full list.

Bias is the difference between the true value and the expected value of the estimator. A specific estimator can be determined to be unbiased or nearly unbiased on the basis of sampling theory for the estimator and the sampling process, but the bias cannot be estimated explicitly from a sample.

Reliability is a measure of the reproducibility of the estimate from the sample over repeated application of the sampling process with the same sample size from the same list. The sampling variance is a commonly used measure of an estimate's reliability. Unbiased estimation of the sampling variance requires that (a) every unit in the list has a known, positive chance of being selected (i.e. the unit selection probability is greater than zero) and (b) every pair of units has a positive chance of being in a sample (the joint selection probability for any pair of units is greater than zero).

List sampling can be performed incorporating a number of different sampling processes depending on the analysis objectives for the sample, the information available for each unit in the list, and data collection procedures. As examples, list sampling can be configured as either single-stage or a multi-stage sampling and with or without stratification.