Optimal experimental design for another’s analysis

Ruth Etzioni, Joseph B. Kadane

Research output: Contribution to journalArticle

36 Citations (Scopus)

Abstract

We consider the optimal design of experiments in which estimation and design are performed by different parties. The parties are assumed to share similar goals, as reflected by a common loss function, but they may have different prior beliefs. After presenting a few motivating examples, we examine the problem of optimal sample size selection under a normal likelihood with constant cost per observation. We also consider the problem of optimal allocation for given overall sample sizes. We present results under both squared-error loss and a logarithmic utility, paying attention to the differences between one- and two-prior optimal designs. An asymmetric discrepancy measure features repeatedly in our development, and we question the extent of its role in optimal two-prior design.

Original languageEnglish (US)
Pages (from-to)1404-1411
Number of pages8
JournalJournal of the American Statistical Association
Volume88
Issue number424
DOIs
StatePublished - Jan 1 1993
Externally publishedYes

Fingerprint

Optimal Experimental Design
Sample Size
Squared Error Loss
Design of Experiments
Optimal Allocation
Loss Function
Discrepancy
Likelihood
Logarithmic
Costs
Design
Experimental design
Sample size
Beliefs
Observation
Loss function
Optimal allocation
Design of experiments

Keywords

  • Adversarial design
  • Bayesian analysis
  • Conjugate prior
  • Optimal allocation
  • Posterior distribution
  • Predictive distribution

ASJC Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

Cite this

Optimal experimental design for another’s analysis. / Etzioni, Ruth; Kadane, Joseph B.

In: Journal of the American Statistical Association, Vol. 88, No. 424, 01.01.1993, p. 1404-1411.

Research output: Contribution to journalArticle

Etzioni, Ruth ; Kadane, Joseph B. / Optimal experimental design for another’s analysis. In: Journal of the American Statistical Association. 1993 ; Vol. 88, No. 424. pp. 1404-1411.
@article{21d4c18016c744ab86434bb65cc89629,
title = "Optimal experimental design for another’s analysis",
abstract = "We consider the optimal design of experiments in which estimation and design are performed by different parties. The parties are assumed to share similar goals, as reflected by a common loss function, but they may have different prior beliefs. After presenting a few motivating examples, we examine the problem of optimal sample size selection under a normal likelihood with constant cost per observation. We also consider the problem of optimal allocation for given overall sample sizes. We present results under both squared-error loss and a logarithmic utility, paying attention to the differences between one- and two-prior optimal designs. An asymmetric discrepancy measure features repeatedly in our development, and we question the extent of its role in optimal two-prior design.",
keywords = "Adversarial design, Bayesian analysis, Conjugate prior, Optimal allocation, Posterior distribution, Predictive distribution",
author = "Ruth Etzioni and Kadane, {Joseph B.}",
year = "1993",
month = "1",
day = "1",
doi = "10.1080/01621459.1993.10476425",
language = "English (US)",
volume = "88",
pages = "1404--1411",
journal = "Journal of the American Statistical Association",
issn = "0162-1459",
publisher = "Taylor and Francis Ltd.",
number = "424",

}

TY - JOUR

T1 - Optimal experimental design for another’s analysis

AU - Etzioni, Ruth

AU - Kadane, Joseph B.

PY - 1993/1/1

Y1 - 1993/1/1

N2 - We consider the optimal design of experiments in which estimation and design are performed by different parties. The parties are assumed to share similar goals, as reflected by a common loss function, but they may have different prior beliefs. After presenting a few motivating examples, we examine the problem of optimal sample size selection under a normal likelihood with constant cost per observation. We also consider the problem of optimal allocation for given overall sample sizes. We present results under both squared-error loss and a logarithmic utility, paying attention to the differences between one- and two-prior optimal designs. An asymmetric discrepancy measure features repeatedly in our development, and we question the extent of its role in optimal two-prior design.

AB - We consider the optimal design of experiments in which estimation and design are performed by different parties. The parties are assumed to share similar goals, as reflected by a common loss function, but they may have different prior beliefs. After presenting a few motivating examples, we examine the problem of optimal sample size selection under a normal likelihood with constant cost per observation. We also consider the problem of optimal allocation for given overall sample sizes. We present results under both squared-error loss and a logarithmic utility, paying attention to the differences between one- and two-prior optimal designs. An asymmetric discrepancy measure features repeatedly in our development, and we question the extent of its role in optimal two-prior design.

KW - Adversarial design

KW - Bayesian analysis

KW - Conjugate prior

KW - Optimal allocation

KW - Posterior distribution

KW - Predictive distribution

UR - http://www.scopus.com/inward/record.url?scp=84950450307&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84950450307&partnerID=8YFLogxK

U2 - 10.1080/01621459.1993.10476425

DO - 10.1080/01621459.1993.10476425

M3 - Article

AN - SCOPUS:84950450307

VL - 88

SP - 1404

EP - 1411

JO - Journal of the American Statistical Association

JF - Journal of the American Statistical Association

SN - 0162-1459

IS - 424

ER -