An empirical evaluation of various priors in the empirical Bayes estimation of small area disease risks

Yutaka Yasui, Hao Liu, Joan Benach, Marcy Winget

Research output: Contribution to journalArticle

13 Citations (Scopus)

Abstract

Empirical and fully Bayes estimation of small area disease risks places a prior distribution on area-specific risks. Several forms of priors have been used for this purpose including gamma, log-normal and non-parametric priors. Spatial correlation among area-specific risks can be incorporated in log-normal priors using Gaussian Markov random fields or other models of spatial dependence. However, the criterion for choosing one prior over others has been mostly logical reasoning. In this paper, we evaluate empirically the various priors used in the empirical Bayes estimation of small area disease risks. We utilize a Spanish mortality data set of a 12-year period to give the underlying true risks, and estimate the true risks using only a 3-year portion of the data set. Empirical Bayes estimates are shown to have substantially smaller mean squared errors than Poisson likelihood-based estimates. However, relative performances of various priors differ across a variety of mortality outcomes considered. In general, the non-parametric prior provides good estimates for lower-risk areas, while spatial priors provide good estimates for higher-risk areas. Ad hoc composite estimates averaging the estimates from the non-parametric prior and those from a spatial log-normal prior appear to perform well overall. This suggests that an empirical Bayes prior that strikes a balance between these two priors, if one can construct such a prior, may prove to be useful for the estimation of small area disease risks. Copyright (C) 2000 John Wiley and Sons, Ltd.

Original languageEnglish (US)
Pages (from-to)2409-2420
Number of pages12
JournalStatistics in Medicine
Volume19
Issue number17-18
StatePublished - Sep 15 2000
Externally publishedYes

Fingerprint

Empirical Bayes Estimation
Evaluation
Estimate
Empirical Bayes
Mortality
Gaussian Markov Random Field
Bayes Estimation
Bayes Estimate
Spatial Dependence
Spatial Correlation
Prior distribution
Mean Squared Error
Averaging
Likelihood
Siméon Denis Poisson
Reasoning
Composite

ASJC Scopus subject areas

  • Epidemiology
  • Statistics and Probability

Cite this

An empirical evaluation of various priors in the empirical Bayes estimation of small area disease risks. / Yasui, Yutaka; Liu, Hao; Benach, Joan; Winget, Marcy.

In: Statistics in Medicine, Vol. 19, No. 17-18, 15.09.2000, p. 2409-2420.

Research output: Contribution to journalArticle

Yasui, Yutaka ; Liu, Hao ; Benach, Joan ; Winget, Marcy. / An empirical evaluation of various priors in the empirical Bayes estimation of small area disease risks. In: Statistics in Medicine. 2000 ; Vol. 19, No. 17-18. pp. 2409-2420.
@article{bb0f3aab705b4f76a8041959db2c8703,
title = "An empirical evaluation of various priors in the empirical Bayes estimation of small area disease risks",
abstract = "Empirical and fully Bayes estimation of small area disease risks places a prior distribution on area-specific risks. Several forms of priors have been used for this purpose including gamma, log-normal and non-parametric priors. Spatial correlation among area-specific risks can be incorporated in log-normal priors using Gaussian Markov random fields or other models of spatial dependence. However, the criterion for choosing one prior over others has been mostly logical reasoning. In this paper, we evaluate empirically the various priors used in the empirical Bayes estimation of small area disease risks. We utilize a Spanish mortality data set of a 12-year period to give the underlying true risks, and estimate the true risks using only a 3-year portion of the data set. Empirical Bayes estimates are shown to have substantially smaller mean squared errors than Poisson likelihood-based estimates. However, relative performances of various priors differ across a variety of mortality outcomes considered. In general, the non-parametric prior provides good estimates for lower-risk areas, while spatial priors provide good estimates for higher-risk areas. Ad hoc composite estimates averaging the estimates from the non-parametric prior and those from a spatial log-normal prior appear to perform well overall. This suggests that an empirical Bayes prior that strikes a balance between these two priors, if one can construct such a prior, may prove to be useful for the estimation of small area disease risks. Copyright (C) 2000 John Wiley and Sons, Ltd.",
author = "Yutaka Yasui and Hao Liu and Joan Benach and Marcy Winget",
year = "2000",
month = "9",
day = "15",
language = "English (US)",
volume = "19",
pages = "2409--2420",
journal = "Statistics in Medicine",
issn = "0277-6715",
publisher = "John Wiley and Sons Ltd",
number = "17-18",

}

TY - JOUR

T1 - An empirical evaluation of various priors in the empirical Bayes estimation of small area disease risks

AU - Yasui, Yutaka

AU - Liu, Hao

AU - Benach, Joan

AU - Winget, Marcy

PY - 2000/9/15

Y1 - 2000/9/15

N2 - Empirical and fully Bayes estimation of small area disease risks places a prior distribution on area-specific risks. Several forms of priors have been used for this purpose including gamma, log-normal and non-parametric priors. Spatial correlation among area-specific risks can be incorporated in log-normal priors using Gaussian Markov random fields or other models of spatial dependence. However, the criterion for choosing one prior over others has been mostly logical reasoning. In this paper, we evaluate empirically the various priors used in the empirical Bayes estimation of small area disease risks. We utilize a Spanish mortality data set of a 12-year period to give the underlying true risks, and estimate the true risks using only a 3-year portion of the data set. Empirical Bayes estimates are shown to have substantially smaller mean squared errors than Poisson likelihood-based estimates. However, relative performances of various priors differ across a variety of mortality outcomes considered. In general, the non-parametric prior provides good estimates for lower-risk areas, while spatial priors provide good estimates for higher-risk areas. Ad hoc composite estimates averaging the estimates from the non-parametric prior and those from a spatial log-normal prior appear to perform well overall. This suggests that an empirical Bayes prior that strikes a balance between these two priors, if one can construct such a prior, may prove to be useful for the estimation of small area disease risks. Copyright (C) 2000 John Wiley and Sons, Ltd.

AB - Empirical and fully Bayes estimation of small area disease risks places a prior distribution on area-specific risks. Several forms of priors have been used for this purpose including gamma, log-normal and non-parametric priors. Spatial correlation among area-specific risks can be incorporated in log-normal priors using Gaussian Markov random fields or other models of spatial dependence. However, the criterion for choosing one prior over others has been mostly logical reasoning. In this paper, we evaluate empirically the various priors used in the empirical Bayes estimation of small area disease risks. We utilize a Spanish mortality data set of a 12-year period to give the underlying true risks, and estimate the true risks using only a 3-year portion of the data set. Empirical Bayes estimates are shown to have substantially smaller mean squared errors than Poisson likelihood-based estimates. However, relative performances of various priors differ across a variety of mortality outcomes considered. In general, the non-parametric prior provides good estimates for lower-risk areas, while spatial priors provide good estimates for higher-risk areas. Ad hoc composite estimates averaging the estimates from the non-parametric prior and those from a spatial log-normal prior appear to perform well overall. This suggests that an empirical Bayes prior that strikes a balance between these two priors, if one can construct such a prior, may prove to be useful for the estimation of small area disease risks. Copyright (C) 2000 John Wiley and Sons, Ltd.

UR - http://www.scopus.com/inward/record.url?scp=0034665661&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0034665661&partnerID=8YFLogxK

M3 - Article

C2 - 10960862

AN - SCOPUS:0034665661

VL - 19

SP - 2409

EP - 2420

JO - Statistics in Medicine

JF - Statistics in Medicine

SN - 0277-6715

IS - 17-18

ER -