A List Sequential Sampling Method Suitable for Real‐Time Sampling

Abstract

A flexible list sequential "π" ps sampling method is introduced and studied. It can reproduce any given sampling design without replacement, of fixed or random sample size. The method is a splitting method and uses successive updating of inclusion probabilities. The main advantage of the method is in real-time sampling situations where it can be used as a powerful alternative to Bernoulli and Poisson sampling and can give any desired second-order inclusion probabilities and thus considerably reduce the variability of the sample size. Copyright (c) Board of the Foundation of the Scandinavian Journal of Statistics 2008.

Spatially correlated Poisson sampling

“…Correlated Poisson Sampling was introduced by Bondesson & Thorburn (2008) as a method suitable for real time sampling with unequal inclusion probabilities. In real time sampling the units of the population are visited by the sampler one by one in some order.…”

“…In fact, every without replacement design with prescribed inclusion probabilities can be implemented by CPS, cf. Bondesson & Thorburn (2008). The method is list sequential, i.e.…”

“…In this paper, the focus is on positive weights. Bondesson & Thorburn (2008) gave some examples on how to choose the weights to get some specific designs. They also derived a general expression for the weights needed to implement any specific design.…”

“…Another important result about the weights, given by Bondesson & Thorburn (2008), is that if the weights w (i) j , j < i, sum to 1 for each j < N , then a fixed sample size is obtained if the π i s sum to an integer. Since a fixed sample size often is desirable, we try to construct weights that sum to 1.…”

“…Such information can be included in the sampling design to make field inventories more efficient. The proposed method is called Spatially Correlated Poisson Sampling (SCPS) and is a modification of Correlated Poisson Sampling (CPS), introduced by Bondesson & Thorburn (2008). With SCPS it is possible to select samples that are spatially balanced.…”