Sampling from a finite population
Citations
1,515 citations
404 citations
242 citations
191 citations
162 citations
Cites methods from "Sampling from a finite population"
...Also the conditional Poisson (CP) design (Hájek, 1981; Tillé, 2006, Chapter 5) is included as a reference to compare against the other methods....
[...]
...Hájek (1981) and Rosén (1997a, 1997b). An even better option is to use a local mean variance estimator, such as the one derived by Stevens and Olsen (2003). It seemed to produce good variance estimates for the GRTS method....
[...]
...Also the conditional Poisson (CP) design (Hájek, 1981; Tillé, 2006, Chapter 5) is included as a reference to compare against the other methods. When the inclusion probabilities are equal, the CP design corresponds to SRS. The CP design totally ignores the spatial aspect because the design is unaffected by a relocation of units within the population. See Section 6 for implementation details. Spatial balance can be measured in different ways. We will use the approach of Voronoi polygons suggested by Stevens and Olsen (2004). We assume that n = ∑ i∈U πi is a positive integer....
[...]
...…entropy and a simple approximative variance estimator, using only first-order inclusion probabilities is the Hájek–Rosén estimator V̂HR(Ŷ ) = n n − 1 ∑ i∈s (1 − πi ) ( yi πi − ∑ j∈s yj (1 − πj )/πj∑ j∈s (1 − πj ) )2 , (5) where s is the sample, cf. Hájek (1981) and Rosén (1997a, 1997b)....
[...]