Stokes (1980) give an estimate of the variance(σ2) under balance ranked set sampling (RSS), and it is proved that this is a gradual unbiased estimates, it is better and more effective than sample variance under simple random sampling (SRS) which the estimation is unbiased. Steven N and Gregory V (2002) give a new variance estimation under balanced ranked set sampling, the estimation is unbiased and estimate is more effective than Stokes’s, without the sorting is perfect. But this article take Steven N and Gregory V (2002)’s unbiased estimation proposed to non-equilibrium ranked set sampling of the case, the resulting estimates are not affected by sorting is perfect or not. |