Diagonal circular systematic sampling

A Note on Diagonal Circular Systematic Sampling

“…The results of efficiency comparisons obtained in Sampath and Varalakshmi (2008) for different values of N , n and k as presented in Tables 1 and 2 seem to be not true. As both CSS and DCSS are not applicable in case (iv) therefore one way to avoid the coincidence of units in the sample is by taking the sampling interval k as the greatest integer smaller thane n N / and use CSS .…”

“…It is observed that some results of efficiency comparisons by Sampath and Varalakshmi (2008) However, in such a case where nk N = , the efficiency comparison using LSS and DSS is quit suitable than using DCSS and CSS. The other comparisons where either CSS or DCSS is not applicable are presented in Tables 1 and 2 …”

“…Starting with a randomly selected element from the N elements of the first row, we choose n units in a diagonal manner as indicated in Example 1 above by following the direction of the arrows. Sampath and varalakshmi (2008) also compared the efficiencies of DCSS and CSS for various combinations of N and n for different natural and simulated populations and conclude that their suggested method performs better than CSS.…”

“…Summary of different combinations of N , n and k where CSS is not applicable but DCSS is.Note: The different combinations of N , n and k presented inTable.2 are those combinations for whichSampath and Varalakshmi (2008) have compared the efficiencies of DCSS and CSS for simulated populations. But by applying Theorem 1, for each possible combination of N , n CSS cannot be applied without coincidence of the units and by DCSS can be applied.…”

Generalized Systematic Sampling

“…This scheme is more general than LSS because it can be applied in both situations whether population size is or not a multiple of the sample size. Subramani (2000) introduced Diagonal Systematic Sampling (DSS) as an alternative to the LSS by assuming that population size N is a multiple of the sample size n, that is, N = nk and n ≤ k. Sampath and Varalakshmi (2008) handled one of the limitations of DSS and suggested Diagonal Circular Systematic Sampling (DCSS) for the case where N = nk. Latter Subramani (2009) relaxed the condition n ≤ k and proposed Generalized Diagonal Systematic Sampling (GDSS), in which sample size n may take a value greater than k. A detailed study of GDSS has been carried out by Subramani (2010).…”