On The Berry-Esseen Theorem for Simple Linear Rank Statistics.

Abstract

The rate of convergence 0(N to the power -1/2 + delta) for any delta > 0 is established for two theorems on asymptotic normality of simple linear rank statistics. These pertain to smooth and bounded scores, arbitrary regression constants, and broad conditions on the distributions of individual observations. The present development provides some alternative arguments of proof, and provides explicit application to relax the conditions of a theorem givin the above rate for the case of location-shift alternatives.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Aug 01, 1977
Accession Number
ADA052320

Entities

People

  • Robert Serfling

Organizations

  • Florida State University

Tags

Communities of Interest

  • Counter IED

DTIC Thesaurus Topics

  • Abstracts
  • Air Force
  • Asymptotic Normality
  • Computing-Related Activities
  • Convergence
  • Data Science
  • Distribution Functions
  • Inequalities
  • Information Science
  • Interdisciplinary Science
  • Military Research
  • Normality
  • Probability
  • Random Variables
  • Statistics
  • United States
  • United States Government

Fields of Study

  • Mathematics

Readers

  • Analytical Mechanics
  • Regression Analysis.