A beta-binomial model for estimating the size of a heterogeneous population

Paul S F Yip, X. I. Liqun, Richard Arnold, Yu Hayakawa

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

This paper compares the properties of various estimators for a beta-binomial model for estimating the size of a heterogeneous population. It is found that maximum likelihood and conditional maximum likelihood estimators perform well for a large population with a large capture proportion. The jackknife and the sample coverage estimators are biased for low capture probabilities. The performance of the martingale estimator is satisfactory, but it requires full capture histories. The Gibbs sampler and Metropolis-Hastings algorithm provide reasonable posterior estimates for informative priors.

Original languageEnglish
Pages (from-to)299-308
Number of pages10
JournalAustralian and New Zealand Journal of Statistics
Volume47
Issue number3
DOIs
Publication statusPublished - 2005 Sep
Externally publishedYes

Fingerprint

Beta-binomial Model
Estimator
Conditional Maximum Likelihood
Metropolis-Hastings Algorithm
Jackknife
Gibbs Sampler
Martingale
Maximum Likelihood Estimator
Biased
Maximum Likelihood
Coverage
Proportion
Estimate

Keywords

  • Beta-binomial
  • Capture-recapture
  • Conditional maximum likelihood estimate
  • Gibbs sampler
  • Maximum likelihood estimate
  • Metropolis-Hastings algorithm

ASJC Scopus subject areas

  • Statistics and Probability

Cite this

A beta-binomial model for estimating the size of a heterogeneous population. / Yip, Paul S F; Liqun, X. I.; Arnold, Richard; Hayakawa, Yu.

In: Australian and New Zealand Journal of Statistics, Vol. 47, No. 3, 09.2005, p. 299-308.

Research output: Contribution to journalArticle

Yip, Paul S F ; Liqun, X. I. ; Arnold, Richard ; Hayakawa, Yu. / A beta-binomial model for estimating the size of a heterogeneous population. In: Australian and New Zealand Journal of Statistics. 2005 ; Vol. 47, No. 3. pp. 299-308.
@article{d4619cf2f5c04d628e47fbdce9c9c09b,
title = "A beta-binomial model for estimating the size of a heterogeneous population",
abstract = "This paper compares the properties of various estimators for a beta-binomial model for estimating the size of a heterogeneous population. It is found that maximum likelihood and conditional maximum likelihood estimators perform well for a large population with a large capture proportion. The jackknife and the sample coverage estimators are biased for low capture probabilities. The performance of the martingale estimator is satisfactory, but it requires full capture histories. The Gibbs sampler and Metropolis-Hastings algorithm provide reasonable posterior estimates for informative priors.",
keywords = "Beta-binomial, Capture-recapture, Conditional maximum likelihood estimate, Gibbs sampler, Maximum likelihood estimate, Metropolis-Hastings algorithm",
author = "Yip, {Paul S F} and Liqun, {X. I.} and Richard Arnold and Yu Hayakawa",
year = "2005",
month = "9",
doi = "10.1111/j.1467-842X.2005.00395.x",
language = "English",
volume = "47",
pages = "299--308",
journal = "Australian and New Zealand Journal of Statistics",
issn = "1369-1473",
publisher = "Wiley-Blackwell",
number = "3",

}

TY - JOUR

T1 - A beta-binomial model for estimating the size of a heterogeneous population

AU - Yip, Paul S F

AU - Liqun, X. I.

AU - Arnold, Richard

AU - Hayakawa, Yu

PY - 2005/9

Y1 - 2005/9

N2 - This paper compares the properties of various estimators for a beta-binomial model for estimating the size of a heterogeneous population. It is found that maximum likelihood and conditional maximum likelihood estimators perform well for a large population with a large capture proportion. The jackknife and the sample coverage estimators are biased for low capture probabilities. The performance of the martingale estimator is satisfactory, but it requires full capture histories. The Gibbs sampler and Metropolis-Hastings algorithm provide reasonable posterior estimates for informative priors.

AB - This paper compares the properties of various estimators for a beta-binomial model for estimating the size of a heterogeneous population. It is found that maximum likelihood and conditional maximum likelihood estimators perform well for a large population with a large capture proportion. The jackknife and the sample coverage estimators are biased for low capture probabilities. The performance of the martingale estimator is satisfactory, but it requires full capture histories. The Gibbs sampler and Metropolis-Hastings algorithm provide reasonable posterior estimates for informative priors.

KW - Beta-binomial

KW - Capture-recapture

KW - Conditional maximum likelihood estimate

KW - Gibbs sampler

KW - Maximum likelihood estimate

KW - Metropolis-Hastings algorithm

UR - http://www.scopus.com/inward/record.url?scp=26244438506&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=26244438506&partnerID=8YFLogxK

U2 - 10.1111/j.1467-842X.2005.00395.x

DO - 10.1111/j.1467-842X.2005.00395.x

M3 - Article

AN - SCOPUS:26244438506

VL - 47

SP - 299

EP - 308

JO - Australian and New Zealand Journal of Statistics

JF - Australian and New Zealand Journal of Statistics

SN - 1369-1473

IS - 3

ER -