Population moments of sampling distributions (Q2488429)

A computer technique is described which yields asymptotic expansions for theoretical moments of sampling distributions. It can be applied to statistics which are functions from empirical moments of univariate or multivariate i.i.d. observations (e.g., for the Pearson correlation). The technique is based on the representation of empirical moments as sums of independent powers, (SIP) \(=\) Kendall's augmented symmetric functions. Implementation of this technique in the Mathematica software is described.