Abstract
We consider the task of Bayesian inference of the mean of normal observations when the available data have been discretized and when no prior knowledge about the mean and the variance exists. An application is presented which illustrates that the discretization of the data should not be ignored when their variability is of the order of the discretization step. We show that the standard (noninformative) prior for location-scale family distributions is no longer appropriate. We work out the reference prior of Berger and Bernardo, which leads to different and more reasonable results. However, for this prior the posterior also shows some non-desirable properties. We argue that this is due to the inherent difficulty of the considered problem, which also affects other methods of inference. We therefore complement our analysis by an empirical Bayes approach. While such proceeding overcomes the disadvantages of the standard and reference priors and appears to provide a reasonable inference, it may raise conceptual concerns. We conclude that it is difficult to provide a widely accepted prior for the considered problem.
Similar content being viewed by others
References
Arnold BC, Press SJ (1986) Bayesian analysis of censored or grouped data from Pareto populations. In: Goel PK, Zellner A (eds) Bayesian inference and decision techniques: essays in honor of Bruno de Finetti (studies in Bayesian econometrics and statistics, vol 6). North-Holland, Amsterdam, pp 157–173
Berger JO (1985) Statistical decision theory and Bayesian analysis. Springer, New York
Berger JO, Bernardo JM (1992) Ordered group reference priors with application to the multinomial problem. Biometrika 79: 25–37
Berger JO, Bernardo JM, Sun D (2009) The formal definition of reference priors. The Ann Stat 37: 905–938
Bernardo JM (2005) Reference analysis. In: Dey DK, Rao CR (eds) Handbook of statistics 25: Bayesian thinking: modeling and computation. Elsevier, Amsterdam, pp 17–90
BIPM, IEC, IFCC, ISO, IUPAC, IUPAP, OIML (1995) Guide to the expression of uncertainty in measurement. Geneva, Switzerland: International Organization for Standardization ISBN 92-67-10188-9
BIPM, IEC, IFCC, ILAC, ISO, IUPAC, IUPAP, and OIML (2008) Evaluation of measurement data—Supplement 1 to the guide to the expression of uncertainty in measurement—propagation of distributions using a monte carlo method. Joint Committee for Guides in Metrology, Bureau International des Poids et Mesures, JCGM, 101: 2008
Cordero RR, Seckmeyer G, Labbe F (2006) Effect of the resolution on the uncertainty evaluation. Metrologia 43: L33–L38
Datta GS, Sweeting TJ (2005) Probability matching priors. In: Dey DK, Rao CR (eds) Handbook of statistics 25: Bayesian thinking: modeling and computation. North-Holland, Amsterdam, pp 91–114
Elster C (2000) Evaluation of measurement uncertainty in the presence of combined random and analogue-to-digital conversion errors. Meas Sci Technol 11: 1359–1363
Frenkel RB (2009) Fiducial inference applied to uncertainty estimation when identical readings are obtained under low instrument resolution. Metrologia 46: 661–667
Frenkel RB, Kirkup L (2005) Monte Carlo-based estimation of uncertainty owing to limited resolution of digital instruments. Metrologia 42: L27–L30
Gleser LJ (1998) Assessing uncertainty in measurement. Stat Sci 13: 277–290
Hannig J, Iyer HK, Wang CM (2007) Fiducial approach to uncertainty assessment accounting for error due to instrument resolution. Metrologia 44: 476–483
Heitjan D (1989) Inference from grouped continuous data: a review. Stat Sci 4: 164–179
Kass RE, Wasserman L (1996) The selection of prior distributions by formal rules. J Am Stat Assoc 91: 1343–1370
Lee C-S, Vardeman SB (2001) Interval estimation of a normal process mean from rounded data. J Qual Technol 33: 335–348
Lee C-S, Vardeman SB (2002) Interval estimation of a normal process standard deviation from rounded data. Commun Stat Simul Comput 31: 13–34
Lira I (2006) Resolution revisited. Metrologia 43: L14–L17
Lira I, Wöger W (1997) The evaluation of standard uncertainty in the presence of limited resolution of indicating devices. Meas Sci Technol 8: 441–443
Locci N, Muscas C, Ghiani E (2002) Evaluation of uncertainty in measurements based on digitized data. Measurement 32: 265–272
Phillips SD, Toman B, Estler WT (2008) Uncertainty due to finite resolution measurements. J Res Natl Inst Stand Technol 113: 143–156
Taraldsen G (2006) Instrument resolution and measurement accuracy. Metrologia 43: 539–544
Vardeman SB, Lee C-S (2005) Likelihood-based statistical estimation from quantized data. IEEE Trans Instrum Meas 54: 409–414
Willink R (2007) On the uncertainty of the mean of discretized measurements. Metrologia 44: 73–81
Witkovsky V, Wimmer G (2009) Interval estimation of the mean of a normal distribution based on quantized observations. Math Slovaca 59: 627–645
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Elster, C., Lira, I. On the choice of a noninformative prior for Bayesian inference of discretized normal observations. Comput Stat 27, 219–235 (2012). https://doi.org/10.1007/s00180-011-0251-7
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00180-011-0251-7