Systematic Sampling Methods

Systematic sampling is a sampling technique that is used for its simplicity and convenience. At its simplest, a systematic sample is obtained by selecting a random start near the beginning of the population list and then taking every unit equally spaced thereafter. The technique can be generalized to include systematic sampling schemes with probability proportional to an auxiliary variable. In many situations, systematic sampling is statistically efficient when compared to other sampling schemes. This is especially true in populations with linear and quadratic trends and autocorrelated populations. For populations in random order systematic sampling, it is equivalent to simple random sampling. A major drawback to systematic sampling is that it does not admit an unbiased estimator of variance with respect to the sampling design. Variance estimation must rely on an assumed underlying structure in the population. Keywords: autocorrelation; linear trend; pps sampling; random order; variance estimation