“School of Mathematics”
Back to Papers HomeBack to Papers of School of Mathematics
Paper IPM / M / 16931  


Abstract:  
Ranked set sampling (RSS) utilizes auxiliary information on the variable of interest so as to assist the experimenter in acquiring an informative sample from the population. The resulting sample has a stratified structure, and often improves statistical inference with respect to the simple random sample of comparable size. In RSS literature, there are some goodnessoffit tests based on the empirical estimators of the instratum cumulative distribution functions (CDFs). Motivated by the fact that the instratum CDFs in RSS can be expressed as functions of the population CDF, some new tests are developed and their asymptotic properties are explored. An extensive simulation study is performed to evaluate properties of different testing procedures when the parent distribution is normal. It turns out that the proposed tests can be considerably more powerful than their contenders in many situations. An application in the context of fishery is also provided.
Download TeX format 

back to top 