Using partial ranking information in the design of small-sample comparisons

George Runger, Thomas R. Willemain

Research output: Contribution to journalArticle

Abstract

Random assignment of experimental units to treatment and control groups is a conventional device to create unbiased comparisons. However, when sample sizes are small and the units differ considerably, there is a significant risk that randomization will create seriously unbalanced partitions of the units into treatment and control groups. We develop and evaluate an alternative to complete randomization for small-sample comparisons involving ordinal data with partial information on ranks of units. For instance, we might know that, of eight units, Rank(A) < Rank(C), Rank(A) < Rank(E) and Rank(D) < Rank(H). We develop an efficient computational procedure to use such information as the basis for restricted randomization of units to the treatment group. We compare our methods to complete randomization in the context of the Mann-Whitney test. With sufficient ranking information, the restricted randomization results in more powerful comparisons.

Original languageEnglish (US)
Pages (from-to)75-86
Number of pages12
JournalJournal of Statistical Computation and Simulation
Volume54
Issue number1-3
StatePublished - 1996
Externally publishedYes

Fingerprint

Small Sample
Ranking
Partial
Information use
Unit
Restricted Randomization
Randomisation
Mann-Whitney test
Ordinal Data
Partial Information
Design
Small sample
Sample Size
Assignment
Partition
Randomization
Sufficient
Evaluate
Alternatives

Keywords

  • Experimental design
  • Mann-Whitney test
  • Randomization; ranking

ASJC Scopus subject areas

  • Applied Mathematics
  • Modeling and Simulation
  • Statistics and Probability

Cite this

Using partial ranking information in the design of small-sample comparisons. / Runger, George; Willemain, Thomas R.

In: Journal of Statistical Computation and Simulation, Vol. 54, No. 1-3, 1996, p. 75-86.

Research output: Contribution to journalArticle

@article{b2c2ee69ad1249099b6e8fa28827a884,
title = "Using partial ranking information in the design of small-sample comparisons",
abstract = "Random assignment of experimental units to treatment and control groups is a conventional device to create unbiased comparisons. However, when sample sizes are small and the units differ considerably, there is a significant risk that randomization will create seriously unbalanced partitions of the units into treatment and control groups. We develop and evaluate an alternative to complete randomization for small-sample comparisons involving ordinal data with partial information on ranks of units. For instance, we might know that, of eight units, Rank(A) < Rank(C), Rank(A) < Rank(E) and Rank(D) < Rank(H). We develop an efficient computational procedure to use such information as the basis for restricted randomization of units to the treatment group. We compare our methods to complete randomization in the context of the Mann-Whitney test. With sufficient ranking information, the restricted randomization results in more powerful comparisons.",
keywords = "Experimental design, Mann-Whitney test, Randomization; ranking",
author = "George Runger and Willemain, {Thomas R.}",
year = "1996",
language = "English (US)",
volume = "54",
pages = "75--86",
journal = "Journal of Statistical Computation and Simulation",
issn = "0094-9655",
publisher = "Taylor and Francis Ltd.",
number = "1-3",

}

TY - JOUR

T1 - Using partial ranking information in the design of small-sample comparisons

AU - Runger, George

AU - Willemain, Thomas R.

PY - 1996

Y1 - 1996

N2 - Random assignment of experimental units to treatment and control groups is a conventional device to create unbiased comparisons. However, when sample sizes are small and the units differ considerably, there is a significant risk that randomization will create seriously unbalanced partitions of the units into treatment and control groups. We develop and evaluate an alternative to complete randomization for small-sample comparisons involving ordinal data with partial information on ranks of units. For instance, we might know that, of eight units, Rank(A) < Rank(C), Rank(A) < Rank(E) and Rank(D) < Rank(H). We develop an efficient computational procedure to use such information as the basis for restricted randomization of units to the treatment group. We compare our methods to complete randomization in the context of the Mann-Whitney test. With sufficient ranking information, the restricted randomization results in more powerful comparisons.

AB - Random assignment of experimental units to treatment and control groups is a conventional device to create unbiased comparisons. However, when sample sizes are small and the units differ considerably, there is a significant risk that randomization will create seriously unbalanced partitions of the units into treatment and control groups. We develop and evaluate an alternative to complete randomization for small-sample comparisons involving ordinal data with partial information on ranks of units. For instance, we might know that, of eight units, Rank(A) < Rank(C), Rank(A) < Rank(E) and Rank(D) < Rank(H). We develop an efficient computational procedure to use such information as the basis for restricted randomization of units to the treatment group. We compare our methods to complete randomization in the context of the Mann-Whitney test. With sufficient ranking information, the restricted randomization results in more powerful comparisons.

KW - Experimental design

KW - Mann-Whitney test

KW - Randomization; ranking

UR - http://www.scopus.com/inward/record.url?scp=0030485943&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0030485943&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:0030485943

VL - 54

SP - 75

EP - 86

JO - Journal of Statistical Computation and Simulation

JF - Journal of Statistical Computation and Simulation

SN - 0094-9655

IS - 1-3

ER -