Lawrence Joseph
Bayesian statistics


Please Note: Prof. Joseph has retired.
These pages are left up in case they prove
useful, but the pages and software will
no longer be updated. All material and
software is "as is" with no guarantees
of functionality or correctness.
Basic info
Course Outline
1 Tues Jan 9 The Zeros and Ones Tricks and Bayesian Conditional Logistic Regression
  Entering non-standard densities in WinBUGS
WinBUGS programming examples of zero and ones tricks
Paper by Mukherjee on Bayesian Conditional Logistic Regression
WinBUGS Program for Conditional Logistic Regression
2 Tues Jan 16 Clinical Trials I
  Lee and Chu 2012
Brophy and Joseph - Placing trials in context using Bayesian analysis
Speigelhalter et al 1994
Hughes 1993
Fayers et al 1997
Fisher 1999
3 Tues Jan 23 Clinical Trials II
4 Tues Jan 30 Sample size I
  Moore 1999
Joseph 1997 (Binomial)
Joseph 1997 (Normal)
Wang and Gelfand 2002
Zou and Normand 2001
Pallay 2000
Practical issues
Download and instructions for R program to calculate Gamma Parameters from Quantiles
5 Tues Feb 6 Sample size II
6 Tues Feb 13 Missing Data
  Simulated CRP Data Set
Kmetic et al: Example of how to handle non-ignorable missing data
7 Tues Feb 20 Measurement Error
8 Tues Feb 27 Adjusting for Unknown Confounding
  Greenland 2003
Steenland 2004
McCandless et al 2007
McCandless et al 2012
Greenland 2005
9 Tues Mar 6 No Class - Spring Break
10 Tues Mar 13 Diagnostic Testing I
  Joseph et al 1995
Dendukuri et al 2001
Ladouceur et al 2006
Dendukuri et al 2004
Weichenthal et al 2010
Program for one test
Program for two tests
An example of running the R program
WinBUGS program for fixed effects correlated model
WinBUGS program for random effects correlated model
11 Tues Mar 20 Diagnostic Testing II
12 Tues Mar 27 Bayesian Meta-Analysis I
  Smith 1994
Carlin 1992
WinBUGS blocker example
Joseph 2000
Brophy 2003
Download and instructions for R forest plot program
13 Tues Apr 3 Bayesian Meta-Analysis II
14 Tues Apr 10 Summary and Conclusion
  Winkler 2001
Berger article
Top 10 reasons to be Bayesian