Confidence intervals for reliability growth models with small sample sizes

Research output: Contribution to journalArticle

16 Citations (Scopus)

Abstract

Fully Bayesian approaches to analysis can be overly ambitious where there exist realistic limitations on the ability of experts to provide prior distributions for all relevant parameters. This research was motivated by situations where expert judgement exists to support the development of prior distributions describing the number of faults potentially inherent within a design but could not support useful descriptions of the rate at which they would be detected during a reliability-growth test. This paper develops inference properties for a reliability-growth model. The approach assumes a prior distribution for the ultimate number of faults that would be exposed if testing were to continue ad infinitum, but estimates the parameters of the intensity function empirically. A fixed-point iteration procedure to obtain the maximum likelihood estimate is investigated for bias and conditions of existence. The main purpose of this model is to support inference in situations where failure data are few. A procedure for providing statistical confidence intervals is investigated and shown to be suitable for small sample sizes. An application of these techniques is illustrated by an example.
LanguageEnglish
Pages257-262
Number of pages6
JournalIEEE Transactions on Reliability
Volume52
Issue number2
DOIs
Publication statusPublished - Jun 2003

Fingerprint

Maximum likelihood
Testing

Keywords

  • reliability
  • reliability management
  • management theory
  • statistics

Cite this

@article{1949fb4fef634971ac592cf4c9eef5e9,
title = "Confidence intervals for reliability growth models with small sample sizes",
abstract = "Fully Bayesian approaches to analysis can be overly ambitious where there exist realistic limitations on the ability of experts to provide prior distributions for all relevant parameters. This research was motivated by situations where expert judgement exists to support the development of prior distributions describing the number of faults potentially inherent within a design but could not support useful descriptions of the rate at which they would be detected during a reliability-growth test. This paper develops inference properties for a reliability-growth model. The approach assumes a prior distribution for the ultimate number of faults that would be exposed if testing were to continue ad infinitum, but estimates the parameters of the intensity function empirically. A fixed-point iteration procedure to obtain the maximum likelihood estimate is investigated for bias and conditions of existence. The main purpose of this model is to support inference in situations where failure data are few. A procedure for providing statistical confidence intervals is investigated and shown to be suitable for small sample sizes. An application of these techniques is illustrated by an example.",
keywords = "reliability, reliability management, management theory, statistics",
author = "J.L. Quigley and L.A. Walls",
year = "2003",
month = "6",
doi = "10.1109/TR.2003.811865",
language = "English",
volume = "52",
pages = "257--262",
journal = "IEEE Transactions on Reliability",
issn = "0018-9529",
number = "2",

}

Confidence intervals for reliability growth models with small sample sizes. / Quigley, J.L.; Walls, L.A.

In: IEEE Transactions on Reliability, Vol. 52, No. 2, 06.2003, p. 257-262.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Confidence intervals for reliability growth models with small sample sizes

AU - Quigley, J.L.

AU - Walls, L.A.

PY - 2003/6

Y1 - 2003/6

N2 - Fully Bayesian approaches to analysis can be overly ambitious where there exist realistic limitations on the ability of experts to provide prior distributions for all relevant parameters. This research was motivated by situations where expert judgement exists to support the development of prior distributions describing the number of faults potentially inherent within a design but could not support useful descriptions of the rate at which they would be detected during a reliability-growth test. This paper develops inference properties for a reliability-growth model. The approach assumes a prior distribution for the ultimate number of faults that would be exposed if testing were to continue ad infinitum, but estimates the parameters of the intensity function empirically. A fixed-point iteration procedure to obtain the maximum likelihood estimate is investigated for bias and conditions of existence. The main purpose of this model is to support inference in situations where failure data are few. A procedure for providing statistical confidence intervals is investigated and shown to be suitable for small sample sizes. An application of these techniques is illustrated by an example.

AB - Fully Bayesian approaches to analysis can be overly ambitious where there exist realistic limitations on the ability of experts to provide prior distributions for all relevant parameters. This research was motivated by situations where expert judgement exists to support the development of prior distributions describing the number of faults potentially inherent within a design but could not support useful descriptions of the rate at which they would be detected during a reliability-growth test. This paper develops inference properties for a reliability-growth model. The approach assumes a prior distribution for the ultimate number of faults that would be exposed if testing were to continue ad infinitum, but estimates the parameters of the intensity function empirically. A fixed-point iteration procedure to obtain the maximum likelihood estimate is investigated for bias and conditions of existence. The main purpose of this model is to support inference in situations where failure data are few. A procedure for providing statistical confidence intervals is investigated and shown to be suitable for small sample sizes. An application of these techniques is illustrated by an example.

KW - reliability

KW - reliability management

KW - management theory

KW - statistics

U2 - 10.1109/TR.2003.811865

DO - 10.1109/TR.2003.811865

M3 - Article

VL - 52

SP - 257

EP - 262

JO - IEEE Transactions on Reliability

T2 - IEEE Transactions on Reliability

JF - IEEE Transactions on Reliability

SN - 0018-9529

IS - 2

ER -