Projective three-level main effects designs robust to model uncertainty

Pi Wen Tsai, Steven G. Gilmour, Roger Mead

Research output: Contribution to journalArticle

32 Citations (Scopus)

Abstract

This paper is concerned with designing experiments under the assumption that an analysis strategy is used that considers interactions in addition to main effects. A criterion which averages an approximation to As-efficiency over lower-dimensional projections of the design is introduced to compare designs. A columnwise design procedure is used to construct three-level designs for six factors in 18 runs. The projection efficiencies of these designs are explored and compared with designs obtained from the L18 orthogonal array. Results are also given for designs with 14 and 17 runs.

Original languageEnglish
Pages (from-to)467-475
Number of pages9
JournalBiometrika
Volume87
Issue number2
DOIs
Publication statusPublished - 2000 Jan 1

Fingerprint

model uncertainty
Robust Design
Main Effect
Model Uncertainty
Uncertainty
Projection
Orthogonal Array
Design
Robust design
Model uncertainty
Approximation
Interaction
Experiment

Keywords

  • Fractional factorial design
  • Response surface methodology
  • Screening experiment
  • Taguchi methods

ASJC Scopus subject areas

  • Statistics and Probability
  • Mathematics(all)
  • Agricultural and Biological Sciences (miscellaneous)
  • Agricultural and Biological Sciences(all)
  • Statistics, Probability and Uncertainty
  • Applied Mathematics

Cite this

Projective three-level main effects designs robust to model uncertainty. / Tsai, Pi Wen; Gilmour, Steven G.; Mead, Roger.

In: Biometrika, Vol. 87, No. 2, 01.01.2000, p. 467-475.

Research output: Contribution to journalArticle

Tsai, Pi Wen ; Gilmour, Steven G. ; Mead, Roger. / Projective three-level main effects designs robust to model uncertainty. In: Biometrika. 2000 ; Vol. 87, No. 2. pp. 467-475.
@article{76f7c3f2f7e041458da11b8332212bd7,
title = "Projective three-level main effects designs robust to model uncertainty",
abstract = "This paper is concerned with designing experiments under the assumption that an analysis strategy is used that considers interactions in addition to main effects. A criterion which averages an approximation to As-efficiency over lower-dimensional projections of the design is introduced to compare designs. A columnwise design procedure is used to construct three-level designs for six factors in 18 runs. The projection efficiencies of these designs are explored and compared with designs obtained from the L18 orthogonal array. Results are also given for designs with 14 and 17 runs.",
keywords = "Fractional factorial design, Response surface methodology, Screening experiment, Taguchi methods",
author = "Tsai, {Pi Wen} and Gilmour, {Steven G.} and Roger Mead",
year = "2000",
month = "1",
day = "1",
doi = "10.1093/biomet/87.2.467",
language = "English",
volume = "87",
pages = "467--475",
journal = "Biometrika",
issn = "0006-3444",
publisher = "Oxford University Press",
number = "2",

}

TY - JOUR

T1 - Projective three-level main effects designs robust to model uncertainty

AU - Tsai, Pi Wen

AU - Gilmour, Steven G.

AU - Mead, Roger

PY - 2000/1/1

Y1 - 2000/1/1

N2 - This paper is concerned with designing experiments under the assumption that an analysis strategy is used that considers interactions in addition to main effects. A criterion which averages an approximation to As-efficiency over lower-dimensional projections of the design is introduced to compare designs. A columnwise design procedure is used to construct three-level designs for six factors in 18 runs. The projection efficiencies of these designs are explored and compared with designs obtained from the L18 orthogonal array. Results are also given for designs with 14 and 17 runs.

AB - This paper is concerned with designing experiments under the assumption that an analysis strategy is used that considers interactions in addition to main effects. A criterion which averages an approximation to As-efficiency over lower-dimensional projections of the design is introduced to compare designs. A columnwise design procedure is used to construct three-level designs for six factors in 18 runs. The projection efficiencies of these designs are explored and compared with designs obtained from the L18 orthogonal array. Results are also given for designs with 14 and 17 runs.

KW - Fractional factorial design

KW - Response surface methodology

KW - Screening experiment

KW - Taguchi methods

UR - http://www.scopus.com/inward/record.url?scp=3142623256&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=3142623256&partnerID=8YFLogxK

U2 - 10.1093/biomet/87.2.467

DO - 10.1093/biomet/87.2.467

M3 - Article

AN - SCOPUS:3142623256

VL - 87

SP - 467

EP - 475

JO - Biometrika

JF - Biometrika

SN - 0006-3444

IS - 2

ER -