The previous discussion in this paper is broadly applicable to both parametric
and nonparametric methodologies. For example, the general property of unbiased for
the sample mean discussed in Section 4 is not dependent on the assumption of any
particular underlying distribution. (The variance of the RSS mean is, of course,
dependent on the underlying distribution through µ, σ
2
, and the µ(i)'s, i = 1, …, k.) For
the remainder of the paper we concentrate solely on a number of important
nonparametric RSS procedures.