English

Estimates for order statistics in terms of quantiles

Probability 2019-01-23 v1

Abstract

Let X1,,XnX_1, \ldots, X_n be independent non-negative random variables with cumulative distribution functions F1,F2,,FnF_1,F_2,\ldots,F_n, each satisfying certain (rather mild) conditions. We show that the median of kk-th smallest order statistic of the vector (X1,,Xn)(X_1, \ldots, X_n) is equivalent to the quantile of order (k1/2)/n(k-1/2)/n with respect to the averaged distribution F=1ni=1nFiF=\frac{1}{n}\sum_{i=1}^n F_i.

Keywords

Cite

@article{arxiv.1901.07398,
  title  = {Estimates for order statistics in terms of quantiles},
  author = {Alexander E. Litvak and Konstantin Tikhomirov},
  journal= {arXiv preprint arXiv:1901.07398},
  year   = {2019}
}

Comments

A small observation published in Zapiski Nauchnykh Seminarov POMI, 457 (2017), 265--275. arXiv admin note: substantial text overlap with arXiv:1609.02126

R2 v1 2026-06-23T07:18:38.632Z