A default prior distribution for contingency tables with dependent factor levels

Antony Overstall; Ruth King

doi:10.1016/j.stamet.2013.08.007

A default prior distribution for contingency tables with dependent factor levels

Antony Overstall, Ruth King

Research output: Contribution to journal › Article › peer-review

Abstract

A default prior distribution is proposed for the Bayesian analysis of contingency tables. The prior is specified to allow for dependence between levels of the factors. Different dependence structures are considered, including conditional autoregressive and distance correlation structures. To demonstrate the prior distribution, a dataset is considered involving estimating the number of injecting drug users in the eleven National Health Service board regions of Scotland using an incomplete contingency table where the dependence structure relates to geographical regions.

Original language	English
Pages (from-to)	90-99
Number of pages	9
Journal	Statistical Methodology
Volume	16
Issue number	1
Early online date	23 Aug 2013
DOIs	https://doi.org/10.1016/j.stamet.2013.08.007
Publication status	Published - Jan 2014

Keywords

Contingency table
Dependence strusture
Default prior

Access to Document

10.1016/j.stamet.2013.08.007

StatMeth16
© 2014 The Authors. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Final published version, 506 KB

Drug use; crime morbidity and mortality: MRC: Drug Use: Crime Morbidity and Mortality
King, R. (PI)
Medical Research Council
1/09/10 → 13/02/14
Project: Standard

Cite this

@article{c6e0e94f37604ad88203b095c4a8abf8,

title = "A default prior distribution for contingency tables with dependent factor levels",

abstract = "A default prior distribution is proposed for the Bayesian analysis of contingency tables. The prior is specified to allow for dependence between levels of the factors. Different dependence structures are considered, including conditional autoregressive and distance correlation structures. To demonstrate the prior distribution, a dataset is considered involving estimating the number of injecting drug users in the eleven National Health Service board regions of Scotland using an incomplete contingency table where the dependence structure relates to geographical regions.",

keywords = "Contingency table, Dependence strusture, Default prior",

author = "Antony Overstall and Ruth King",

note = "Both authors were partly funded by MRC-funded addictions cluster, NIQUAD (Grant No. G1000021).",

year = "2014",

month = jan,

doi = "10.1016/j.stamet.2013.08.007",

language = "English",

volume = "16",

pages = "90--99",

journal = "Statistical Methodology",

issn = "1572-3127",

publisher = "Elsevier",

number = "1",

}

TY - JOUR

T1 - A default prior distribution for contingency tables with dependent factor levels

AU - Overstall, Antony

AU - King, Ruth

N1 - Both authors were partly funded by MRC-funded addictions cluster, NIQUAD (Grant No. G1000021).

PY - 2014/1

Y1 - 2014/1

N2 - A default prior distribution is proposed for the Bayesian analysis of contingency tables. The prior is specified to allow for dependence between levels of the factors. Different dependence structures are considered, including conditional autoregressive and distance correlation structures. To demonstrate the prior distribution, a dataset is considered involving estimating the number of injecting drug users in the eleven National Health Service board regions of Scotland using an incomplete contingency table where the dependence structure relates to geographical regions.

AB - A default prior distribution is proposed for the Bayesian analysis of contingency tables. The prior is specified to allow for dependence between levels of the factors. Different dependence structures are considered, including conditional autoregressive and distance correlation structures. To demonstrate the prior distribution, a dataset is considered involving estimating the number of injecting drug users in the eleven National Health Service board regions of Scotland using an incomplete contingency table where the dependence structure relates to geographical regions.

KW - Contingency table

KW - Dependence strusture

KW - Default prior

U2 - 10.1016/j.stamet.2013.08.007

DO - 10.1016/j.stamet.2013.08.007

M3 - Article

SN - 1572-3127

VL - 16

SP - 90

EP - 99

JO - Statistical Methodology

JF - Statistical Methodology

IS - 1

ER -

A default prior distribution for contingency tables with dependent factor levels

Abstract

Keywords

Access to Document

Fingerprint

Projects

Drug use; crime morbidity and mortality: MRC: Drug Use: Crime Morbidity and Mortality

Cite this