An intriguing example ("outlying posterior phenomenon" -- now named "discrepant posterior phenomenon")


Posterior is more extreme than both the prior and the likelihood!  <== Is it a paradox or something else?


A review article on confidence distributions and some related manuscripts:         


                        Confidence Distribution, the Frequentist Distribution Estimator of a Parameter -- a Review (International Statistical Review, 2012) [Journal Link]
                               (To appear with discussions by David R. Cox, Brad Efron, Emanuel Parzen, D.A.S. Fraser, Tore Schweder/Nils Hjort and Christian Robert)

                         Related reading (two cited examples --- discrepant (outlying) posterior phenomenon in Section 6.2 and CD-based simulaiton in Section 7.3):
                         Incorporating External Information in Analysis of Clinical Trials with Binary Outcomes (AOAS, 2012)
                         An inference procedure for order parameters utilizing conffidence distribution random variables (Techincal report)
                        

Peer Reviewed Research Papers:


Confidence Distributions and a Unified Framework for Meta-analysis (JASA, 2011) [Journal Link]

Discussions on Professor Fraser’s article on “Is Bayes posterior just quick and dirty confidence?” (Statistical Sciences, 2011) 

CD Posterior - Combining Prior and Data Through Confidence Distributions to apear in "Contemporary Developments in Bayesian Analysis and Statistical Decision Theory : A Festschrift for William E. Strawderman". IMS Collection, Volume 8. (D. Fourdrinier, et al. Eds) (2012).

Incorporating Expert Opinions with Information from Binomial Clinical Trials  (AOAS, 2012)

Sensor Management Problems of Nuclear Detection  (in "Safty and Risk Modeling and its Applications", Springer (H. Pham ed.), 2011).

Investigating the Impact of Uncertainty About Item Parameters on Ability Estimation(Psychometrika 2011). [Journal Link]

Port of entry inspection policies: Incorporation of measurement errors. (Annals of Operational Research, 2011). [Journal Link] 

A Note on Dichotomization of Continuous Response Variable in Presence of Measuring Errors (Statistics in Medicine, 2010). [Journal Link]

Multi-Objective Optimization of a Port-of-Entry Inspection Policy.  (IEEE Transaction-ASE, 2010). [Journal Link]

Bootstrap Method. (International Encyclopedia of Education, 2010). [Publication Link]

Confidence Intervals for Population Ranks in the Presence of Ties or Near Ties. (JASA, 2009) [Journal Link]

A latent model to detecting multiple temporal clusters.(Biometrics, 2009) [Supplementary].

Port-of Entry Inspection: Sensor Deployment Policy and Optimization.  (IEEE Transactions-ASE, 2009).

Patient Complexity in Quality Comparisons for Glycemic Control: an Observational Study.  (Implementation Science, 2009).

Analysis of Heterogeneous Data Using Varying-Scale Generalized Linear Models. (JASA, 2008).  [Supplementary.]  [Journal Link]

Optimization Problems for Port-of-Entry Detection Systems (in Intelligence and Security Informatics: Techniques and Applications, Edited by H. Chen and C. Yang., 2008).

Cluster detection and Pervasive Surveillance of Nuclear Materials in Metropolitan Area Using Taxi Cabs. (DIMACS Tech Report)


Recent Talks:


          Confidence Distribution, the Frequentisst Distribution Estimator of a Parameter & a General Framework for Combining Information

A General Framework for Combining Information & A Frequentist Approach to Incorporate Expert Opinions

A Latent Model to Detect Multiple Clusters of Varying Sizes

Statistical Inference for Ranks of Health Care Facilities in the Presence of Ties and Near Ties


Software Package (to be submitted to CRAN):


gmeta: Meta-Analysis via a Unified Framework under Confidence Distribution