Abstract
ABSTRACT Simple Latin Square Sampling −k (SLSS–k) design was introduced by Borkowski (2003) for a situation that a positive spatial correlation is present in a square population. In this article, we introduce a new sampling design called Row and Column Elimination Sampling (RCES) design which is an extension of the SLSS–k to any rectangle population. There is no unbiased variance estimation for the SLSS–k design and no other variance estimation is available. Not having a variance estimator is a serious weak point of a sampling design from practical prospective. With taking an extra sample unit, we provide variance estimator for the RCES + 1 design. We then introduce adaptive cluster RCES + 1 design. We evaluate estimators of the sampling designs using real life populations and an artificial population. Estimators of adaptive and non adaptive versions of RCES + 1 have smaller variances than their simple random sample counterparts.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
