Regression Efficient Analysis With Ranked Set Sampling

Ranked set sampling (Ranked set sampling short for RSS) in many cases is a more effective way of data collection than simple random sampling. So since the 1990s the statistical inference based on ranked set sampling method has been paid more and more attention. There are many beautiful results on the hypothesis testing, parameter estimation and non parametric statistics so far. Chen and Wang (2004) in the view of the efficiency of regression coefficient estimates build a ranked set sampling model. In this paper, on the basis of this article, in the view of the efficiency of regression coefficient estimates establish a sampling method to estimate one regression and quadratic regression coefficient on the basis of D-optimum as standard. Through data simulation, validated using extreme ranked set sampling and weighting ranked set sampling respectively to a linear regression and quadratic linear regression coefficient estimation efficiency compared with the simple random sampling is improved greatly.