Nonparametric Procedures in Multiple Decisions (Ranking and Selection Procedures)

Abstract

This article surveys statistical techniques which are nonparametric in nature and used in formal ranking and selection of populations. Such methods have been developed only within the last fifteen years and are usually based on rank scores and/or robust estimators (such as the Hodges-Lehmann estimator). The procedures surveyed are applicable to one-way classifications, two-way classifications, and paired-comparison models. Computational methods, useful inequalities, and appropriate numerical tables required to implement these techniques are identified and discussed. Asymptotic relative efficiencies of the nonparametric methods, compared to their parametric counterparts, are presented. Specific applications of these methods (such as traffic fatality rates) are mentioned and areas for further theoretical and computational research are identified.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
May 01, 1980
Accession Number
ADA096093

Entities

People

  • Gary C. Mcdonald
  • Shanti Gupta

Organizations

  • Purdue University

Tags

Communities of Interest

  • Human Systems

DTIC Thesaurus Topics

  • Classification
  • Computational Science
  • Data Science
  • Decision Theory
  • Efficiency
  • Estimators
  • Fatalities
  • Inequalities
  • Information Science
  • New York
  • Observation
  • Probability
  • Random Variables
  • Statistical Algorithms
  • Statistical Inference
  • Statistics
  • Surveys

Fields of Study

  • Mathematics

Readers

  • Regression Analysis.
  • Statistical inference.