Bayesian estimation of a random effects heteroscedastic probit model

Yuanyuan Gu, Denzil G. Fiebig, Edward Cripps, Robert Kohn

Research output: Contribution to journalArticleResearchpeer-review

Abstract

Bayesian analysis is given of a random effects binary probit model that allows for heteroscedasticity. Real and simulated examples illustrate the approach and show that ignoring heteroscedasticity when it exists may lead to biased estimates and poor prediction. The computation is carried out by an efficient Markov chain Monte Carlo sampling scheme that generates the parameters in blocks. We use the Bayes factor, cross-validation of the predictive density, the deviance information criterion and Receiver Operating Characteristic (ROC) curves for model comparison.

LanguageEnglish
Pages324-339
Number of pages16
JournalEconometrics Journal
Volume12
Issue number2
DOIs
Publication statusPublished - 2009
Externally publishedYes

Fingerprint

Random effects probit
Heteroscedasticity
Probit model
Bayesian estimation
Bayesian analysis
Cross-validation
Prediction
Model comparison
Random effects
Predictive density
Markov chain Monte Carlo
Bayes factor
Receiver operating characteristic curve
Sampling
Information criterion
Deviance

Keywords

  • Bayes factor
  • Cross-validation
  • Deviance information criterion
  • Marginal effects
  • Marginal likelihood
  • Markov chain Monte Carlo
  • ROC curve

Cite this

Gu, Yuanyuan ; Fiebig, Denzil G. ; Cripps, Edward ; Kohn, Robert. / Bayesian estimation of a random effects heteroscedastic probit model. In: Econometrics Journal. 2009 ; Vol. 12, No. 2. pp. 324-339.
@article{0409ae077dea4fe1b561ee675dc14c25,
title = "Bayesian estimation of a random effects heteroscedastic probit model",
abstract = "Bayesian analysis is given of a random effects binary probit model that allows for heteroscedasticity. Real and simulated examples illustrate the approach and show that ignoring heteroscedasticity when it exists may lead to biased estimates and poor prediction. The computation is carried out by an efficient Markov chain Monte Carlo sampling scheme that generates the parameters in blocks. We use the Bayes factor, cross-validation of the predictive density, the deviance information criterion and Receiver Operating Characteristic (ROC) curves for model comparison.",
keywords = "Bayes factor, Cross-validation, Deviance information criterion, Marginal effects, Marginal likelihood, Markov chain Monte Carlo, ROC curve",
author = "Yuanyuan Gu and Fiebig, {Denzil G.} and Edward Cripps and Robert Kohn",
year = "2009",
doi = "10.1111/j.1368-423X.2009.00283.x",
language = "English",
volume = "12",
pages = "324--339",
journal = "Econometrics Journal",
issn = "1368-4221",
publisher = "Blackwell",
number = "2",

}

Bayesian estimation of a random effects heteroscedastic probit model. / Gu, Yuanyuan; Fiebig, Denzil G.; Cripps, Edward; Kohn, Robert.

In: Econometrics Journal, Vol. 12, No. 2, 2009, p. 324-339.

Research output: Contribution to journalArticleResearchpeer-review

TY - JOUR

T1 - Bayesian estimation of a random effects heteroscedastic probit model

AU - Gu, Yuanyuan

AU - Fiebig, Denzil G.

AU - Cripps, Edward

AU - Kohn, Robert

PY - 2009

Y1 - 2009

N2 - Bayesian analysis is given of a random effects binary probit model that allows for heteroscedasticity. Real and simulated examples illustrate the approach and show that ignoring heteroscedasticity when it exists may lead to biased estimates and poor prediction. The computation is carried out by an efficient Markov chain Monte Carlo sampling scheme that generates the parameters in blocks. We use the Bayes factor, cross-validation of the predictive density, the deviance information criterion and Receiver Operating Characteristic (ROC) curves for model comparison.

AB - Bayesian analysis is given of a random effects binary probit model that allows for heteroscedasticity. Real and simulated examples illustrate the approach and show that ignoring heteroscedasticity when it exists may lead to biased estimates and poor prediction. The computation is carried out by an efficient Markov chain Monte Carlo sampling scheme that generates the parameters in blocks. We use the Bayes factor, cross-validation of the predictive density, the deviance information criterion and Receiver Operating Characteristic (ROC) curves for model comparison.

KW - Bayes factor

KW - Cross-validation

KW - Deviance information criterion

KW - Marginal effects

KW - Marginal likelihood

KW - Markov chain Monte Carlo

KW - ROC curve

UR - http://www.scopus.com/inward/record.url?scp=68149171594&partnerID=8YFLogxK

U2 - 10.1111/j.1368-423X.2009.00283.x

DO - 10.1111/j.1368-423X.2009.00283.x

M3 - Article

VL - 12

SP - 324

EP - 339

JO - Econometrics Journal

T2 - Econometrics Journal

JF - Econometrics Journal

SN - 1368-4221

IS - 2

ER -