Graphical tools for assessing the sensitivity of response surface designs to model misspecification

Christine M. Anderson-Cook, Connie M. Borror, Bradley Jones

Research output: Contribution to journalArticle

6 Citations (Scopus)

Abstract

When comparing different designs for an experiment, optimality criteria and other measures often depend on the correctness of the assumed model. In this article we develop and illustrate an approach for comparing designs given the potential effect of bias due to an underspecified model. We illustrate this approach using graphical summaries of the expected mean squared error (EMSE) that allow assessment of the robustness of designs to model misspecification. For response surface designs in cuboidal regions, the excellent performance of the central composite designs when the quadratic model is correct is tempered when bias is present. The Box-Behnken designs demonstrate superiority in the presence of missing cubic terms. For a screening design scenario, the amount of bias present leads to different conclusions as to which design is best.

Original languageEnglish (US)
Pages (from-to)75-87
Number of pages13
JournalTechnometrics
Volume51
Issue number1
DOIs
StatePublished - 2009

Fingerprint

Response Surface Design
Model Misspecification
Screening Design
Optimality Criteria
Mean Squared Error
Design
Graphics
Correctness
Composite
Model
Robustness
Scenarios
Screening
Term
Demonstrate
Experiment

Keywords

  • Expected squared bias
  • Fraction of design space plots
  • Mean squared error
  • Prediction variance

ASJC Scopus subject areas

  • Modeling and Simulation
  • Statistics and Probability
  • Applied Mathematics

Cite this

Graphical tools for assessing the sensitivity of response surface designs to model misspecification. / Anderson-Cook, Christine M.; Borror, Connie M.; Jones, Bradley.

In: Technometrics, Vol. 51, No. 1, 2009, p. 75-87.

Research output: Contribution to journalArticle

Anderson-Cook, Christine M. ; Borror, Connie M. ; Jones, Bradley. / Graphical tools for assessing the sensitivity of response surface designs to model misspecification. In: Technometrics. 2009 ; Vol. 51, No. 1. pp. 75-87.
@article{7da3b207cbd244e6b199fb91c44c45a3,
title = "Graphical tools for assessing the sensitivity of response surface designs to model misspecification",
abstract = "When comparing different designs for an experiment, optimality criteria and other measures often depend on the correctness of the assumed model. In this article we develop and illustrate an approach for comparing designs given the potential effect of bias due to an underspecified model. We illustrate this approach using graphical summaries of the expected mean squared error (EMSE) that allow assessment of the robustness of designs to model misspecification. For response surface designs in cuboidal regions, the excellent performance of the central composite designs when the quadratic model is correct is tempered when bias is present. The Box-Behnken designs demonstrate superiority in the presence of missing cubic terms. For a screening design scenario, the amount of bias present leads to different conclusions as to which design is best.",
keywords = "Expected squared bias, Fraction of design space plots, Mean squared error, Prediction variance",
author = "Anderson-Cook, {Christine M.} and Borror, {Connie M.} and Bradley Jones",
year = "2009",
doi = "10.1198/TECH.2009.0008",
language = "English (US)",
volume = "51",
pages = "75--87",
journal = "Technometrics",
issn = "0040-1706",
publisher = "American Statistical Association",
number = "1",

}

TY - JOUR

T1 - Graphical tools for assessing the sensitivity of response surface designs to model misspecification

AU - Anderson-Cook, Christine M.

AU - Borror, Connie M.

AU - Jones, Bradley

PY - 2009

Y1 - 2009

N2 - When comparing different designs for an experiment, optimality criteria and other measures often depend on the correctness of the assumed model. In this article we develop and illustrate an approach for comparing designs given the potential effect of bias due to an underspecified model. We illustrate this approach using graphical summaries of the expected mean squared error (EMSE) that allow assessment of the robustness of designs to model misspecification. For response surface designs in cuboidal regions, the excellent performance of the central composite designs when the quadratic model is correct is tempered when bias is present. The Box-Behnken designs demonstrate superiority in the presence of missing cubic terms. For a screening design scenario, the amount of bias present leads to different conclusions as to which design is best.

AB - When comparing different designs for an experiment, optimality criteria and other measures often depend on the correctness of the assumed model. In this article we develop and illustrate an approach for comparing designs given the potential effect of bias due to an underspecified model. We illustrate this approach using graphical summaries of the expected mean squared error (EMSE) that allow assessment of the robustness of designs to model misspecification. For response surface designs in cuboidal regions, the excellent performance of the central composite designs when the quadratic model is correct is tempered when bias is present. The Box-Behnken designs demonstrate superiority in the presence of missing cubic terms. For a screening design scenario, the amount of bias present leads to different conclusions as to which design is best.

KW - Expected squared bias

KW - Fraction of design space plots

KW - Mean squared error

KW - Prediction variance

UR - http://www.scopus.com/inward/record.url?scp=77950844370&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77950844370&partnerID=8YFLogxK

U2 - 10.1198/TECH.2009.0008

DO - 10.1198/TECH.2009.0008

M3 - Article

AN - SCOPUS:77950844370

VL - 51

SP - 75

EP - 87

JO - Technometrics

JF - Technometrics

SN - 0040-1706

IS - 1

ER -