Interval estimation via tail functions

Borek Puza, Terence O'Neill

Research output: Contribution to journalArticleResearchpeer-review

10 Citations (Scopus)

Abstract

The authors describe a new method for constructing confidence intervals. Their idea consists in specifying the cutoff points in terms of a function of the target parameter rather than as constants. When it is suitably chosen, this so-called tail function yields shorter confidence intervals in the presence of prior information. It can also be used to improve the coverage properties of approximate confidence intervals. The authors illustrate their technique by application to interval estimation of the mean of Bernoulli and normal populations. They further suggest guidelines for choosing the optimal tail function and discuss the relationship with Bayesian inference.

Original languageEnglish
Pages (from-to)299-310
Number of pages12
JournalCanadian Journal of Statistics
Volume34
Issue number2
DOIs
Publication statusPublished - 1 Jun 2006
Externally publishedYes

Fingerprint

Interval Estimation
Confidence interval
Tail
Normal Population
Prior Information
Bayesian inference
Bernoulli
Coverage
Target
Interval estimation

Cite this

Puza, Borek ; O'Neill, Terence. / Interval estimation via tail functions. In: Canadian Journal of Statistics. 2006 ; Vol. 34, No. 2. pp. 299-310.
@article{eb482b0efc06421eaf0adacb01511fca,
title = "Interval estimation via tail functions",
abstract = "The authors describe a new method for constructing confidence intervals. Their idea consists in specifying the cutoff points in terms of a function of the target parameter rather than as constants. When it is suitably chosen, this so-called tail function yields shorter confidence intervals in the presence of prior information. It can also be used to improve the coverage properties of approximate confidence intervals. The authors illustrate their technique by application to interval estimation of the mean of Bernoulli and normal populations. They further suggest guidelines for choosing the optimal tail function and discuss the relationship with Bayesian inference.",
author = "Borek Puza and Terence O'Neill",
year = "2006",
month = "6",
day = "1",
doi = "10.1002/cjs.5550340207",
language = "English",
volume = "34",
pages = "299--310",
journal = "Canadian Journal of Statistics",
issn = "0319-5724",
publisher = "Statistical Society of Canada",
number = "2",

}

Interval estimation via tail functions. / Puza, Borek; O'Neill, Terence.

In: Canadian Journal of Statistics, Vol. 34, No. 2, 01.06.2006, p. 299-310.

Research output: Contribution to journalArticleResearchpeer-review

TY - JOUR

T1 - Interval estimation via tail functions

AU - Puza, Borek

AU - O'Neill, Terence

PY - 2006/6/1

Y1 - 2006/6/1

N2 - The authors describe a new method for constructing confidence intervals. Their idea consists in specifying the cutoff points in terms of a function of the target parameter rather than as constants. When it is suitably chosen, this so-called tail function yields shorter confidence intervals in the presence of prior information. It can also be used to improve the coverage properties of approximate confidence intervals. The authors illustrate their technique by application to interval estimation of the mean of Bernoulli and normal populations. They further suggest guidelines for choosing the optimal tail function and discuss the relationship with Bayesian inference.

AB - The authors describe a new method for constructing confidence intervals. Their idea consists in specifying the cutoff points in terms of a function of the target parameter rather than as constants. When it is suitably chosen, this so-called tail function yields shorter confidence intervals in the presence of prior information. It can also be used to improve the coverage properties of approximate confidence intervals. The authors illustrate their technique by application to interval estimation of the mean of Bernoulli and normal populations. They further suggest guidelines for choosing the optimal tail function and discuss the relationship with Bayesian inference.

UR - http://www.scopus.com/inward/record.url?scp=33748310558&partnerID=8YFLogxK

U2 - 10.1002/cjs.5550340207

DO - 10.1002/cjs.5550340207

M3 - Article

VL - 34

SP - 299

EP - 310

JO - Canadian Journal of Statistics

JF - Canadian Journal of Statistics

SN - 0319-5724

IS - 2

ER -