Abstract
A new scheme ‘Rhombus Ranked Set Sampling’ (RRSS) is developed in this research together with its properties for estimating the population means. Mathematical validation along with the simulation evaluation is presented. The proposed method is an addition to the family of different sampling methods and generalization of ‘Folded Ranked Set Sampling’ (FRSS). For the simulation process, nine probability distributions are considered for the efficiency comparison of proposed scheme from which four are symmetric and rest are asymmetric among which Weibull and beta distributions which are used twice, unlike parametric values. (Al-Naseer, 2007 and Bani-Mustafa, 2011). Through simulation processes, it is observed that RRSS is competent and more reliable relative to simple random sampling (SRS), ranked set sampling (RSS) and folded ranked set sampling (FRSS). It is noted that for all the underlying distributions, an increase in the efficiency of Rhombus Ranked Set Sampling (RRSS) is achieved via increasing the size of the sample ‘p’. Besides the efficiency comparison, consistency of the proposed method is also valued by using Co-efficient of Variation (CV). Secondary data on zinc (Zn) concentration and lead (Pb) contamination in different parts and tissues of freshwater fish was collected to illustrate the evaluation of RRSS against SRS, RSS, FRSS and ERSS (extreme ranked set sampling). The results obtained through real life illustration defend the simulation study and hence indicates that the RRSS estimator is efficient substitute for existing methods (Al-Omari, 2011).
Highlights
‘Ranked set sampling’ (RSS) is a non-parametric methodology of assembling data that improves estimation through sampler’s belief or by using statistics of sampling units (Bohn. 1996; Presnell & Bohn, 1999; Sroka, 2008 and Barabesi & El-Sharaawi, 2001)
A new tool ‘Rhombus Ranked Set Sampling’ (RRSS) is introduced, an extension to ‘Folded Ranked Set Sampling’ (FRSS)’, whose working is healthier than almost all the advancements made in ranked set sampling (RSS) in the preceding years
It is interesting to note that when the observations are observed from their own row, they will be dependent and as a result covariance term exists whereas, the observations will be independent when studied from different rows
Summary
‘Ranked set sampling’ (RSS) is a non-parametric methodology of assembling data that improves estimation through sampler’s belief or by using statistics of sampling units (Bohn. 1996; Presnell & Bohn, 1999; Sroka, 2008 and Barabesi & El-Sharaawi, 2001). ‘Ranked set sampling’ (RSS) is a non-parametric methodology of assembling data that improves estimation through sampler’s belief or by using statistics of sampling units 1996; Presnell & Bohn, 1999; Sroka, 2008 and Barabesi & El-Sharaawi, 2001) It is amongst one of the most accepted and reliable method commenced by McIntyre (1952) that, over the years, have been widely studied for estimations of parameters of different distributions (Muttlak & McDonald, 1992; Fei et al, 1994 and Lam et al, 1994). A further addition named as ‘Folded ranked set sampling’ has been introduced in recent years to overcome the problem of wastage of sampling units (Bani-Mustafa et al, 2011).
Sign-in/Register to view highlights & summary 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
