Detecting the Influential Observation Using Intrinsic Bayes Factors

  • 발행 : 2000.03.01

초록

For the balanced variance component model, sometimes intraclass correlation coefficient is of interest. If there is little information about the parameter, then the reference prior(Berger and Bernardo, 1992) is widely used. Pettit nd Young(1990) considered a measrue of the effect of a single observation on a logarithmic Bayes factor. However, under such a reference prior, the Bayes factor depends on the ratio of unspecified constants. In order to discard this problem, influence diagnostic measures using the intrinsic Bayes factor(Berger and Pericchi, 1996) is presented. Finally, one simulated dataset is provided which illustrates the methodology with appropriate simulation based computational formulas. In order to overcome the difficult Bayesian computation, MCMC methods, such as Gibbs sampler(Gelfand and Smith, 1990) and Metropolis algorithm, are empolyed.

키워드

참고문헌

  1. Journal of the Royal Statistical Society v.53 Posterior Bayes factor (with discussion) Aitkin;M.
  2. Bayesian Statistics v.4 On the development of the reference prior mwthod. Berger;J.;J.;Bernardo;J.M. Bernardo;J.O. Berger;A.P. Dawid;A.F.M. Smith
  3. Journal of the American statistical Association v.91 The intrinsic Bayes factor for model selsction and prediction Berger;J.O.;Pericchi;L.R.
  4. J. Roy. Statis. Soc. v.41 Reference posterior distributions for Bayesian inference(with Discussion) Bernardo;J.
  5. Assison-Wesley Bayesian Inference in Statistical Analysis Box;G.;G. Tiao
  6. Technometrics v.29 A Bayesian approach to the estimation of variance component for the unbalanced one-way rendom model Chaloner;K.
  7. Journal of American Statistical Association v.89 Importance-weighted marginal Bayesian posterior density estimation Chen;M.H.
  8. Communications in Statistics v.27 no.9 Bayesian approach to estimation of intraclass correlation using reference prior Chung;Y.;Dey;D. K.
  9. In: Bayesian Statistics 5(J. Barnardo. et al,eds.) Discussion of intrinsic Bayes factor for linear model Dey;D. K.
  10. In: Bayesian statistics 4 (J. Bernardo,et. al. eds.) Model determination using predictive distributions with impletation via sampling-based mothods Gelfand;A. E.;Dey;D. K.;Chang;H.
  11. Journal of the American Statistical Association v.85 Sampling based approach to calculating marginal densities Gelfand;A. E.;Smith;A.F.M.
  12. Comm. Statist. Theory and Methods v.19 A comparison of Bayes and maximum likelihood estimation of the intraclass clrrelation coefficient Plamer;J.L.;L.D. Broemeling
  13. Journal of the Royal Statistical Society v.52 The conditional presicitive ordinate for the normal distribution Pattit;L.I.
  14. Journal of the Royal Statistical Society v.77 Measuring the effect of observations on Bayes factors Pattit;L.I.;Young;K.D.S.
  15. Journal of American Statistical Association v.90 The reference prior Batesian analysis for normal mean produsts Sun D.;Ye. K.
  16. Journal of Statistical Planning Inference v.41 Bayesian reference prior analysis on the ration of variances for the balanced one-way random effect model Ye;K.