A sharpening of the remainder term in the higher-dimensional central limit theorem for multilinear rank statistics (Q1072280)

A new approach to the asymptotic normality of linear rank statistic is provided by \textit{M. Denker} and \textit{U. Rösler} in Stat. Decis. 3, 49- 75 (1985; Zbl 0568.62017). This method is used to prove the asymptotic normality of multivariate linear rank statistics together with Berry- Esséen and Prokhorov distance estimates. In particular - up to a factor of (log n)\({}^ 2\)- the optimal rates are obtained for score functions \(\phi\) with continuous bounded second derivatives or with continuous derivatives \(\phi\) ' satisfying \(| \phi '(t)| \leq C_ 0(t(1-t))^{-3/2+\eta}\) for some \(C_ 0\), \(\eta >0\).