Lawrence Joseph
Bayesian statistics
Professor
E-mail: lawrence.joseph@mcgill.ca
Telephone: +1 (514) 934-1934 ext: 44713
FAX: (514) 934-8293
Complete CV
Division of Clinical Epidemiology
McGill University Health Centre
Royal Victoria Hospital
687 Pine Avenue West
V Building, Room V2.10
Montreal, Quebec
Canada, H3A 1A1
Course Outline
Introductory Lecture
[1] Jan 8, 2001, one hour only
  • Introduction to the course, including overview and selection of topics, grading scheme, and assignments.
  • Briefly review reading list to be discussed next week.
  • Handout summarizing basic elements used in Bayesian analysis
  • Selected pages from A. Gelman, J. Carlin, H. Stern and D. Rubin, Bayesian Data Analysis, (1995, Chapman and Hall), pages 3 to 10.
  • Selected pages from D. Berry and D. Stangl, Bayesian Biostatistics (1996, Dekker), pages 1 to 20.
  • Selected pages from J. Berger, Statistical decision theory and Bayesian analysis, (1985), Springer Verlag, pages 20-35.
  • J. Berger and D. Berry. Statistical analysis and the illusion of objectivity, (1988), American Scientist, 159--165
Introduction to the Bayesian Approach to Statistical Analysis
[2] Jan 15, 2001

Discussion based on the readings, focussing on the following topics:
  • Review of the correct interpretations of frequentist p-values and confidence intervals
  • Problems with frequentist analysis
  • Bayes Theorem
  • Subjective or objective analyses
  • The likelihood principle
  • Inferential or decision theoretic Bayesian analysis
  • Problems with Bayesian analysis and possible solutions
Conjugate Bayesian Inference for Proportions and Means
[3] Jan 22, 2001
  • Prior distributions (various choices and discussion)
  • One sample problem for one proportion
  • One sample problem for one mean, variance known
  • Selected pages from A. Gelman, J. Carlin, H. Stern and D. Rubin, Bayesian Data Analysis, (1995), pages 28 to 58.
  • Notes from course pack.
Practical Implementation I
[4] Jan 29, 2001
  • Numerical methods, Monte Carlo integration and the Gibbs sampler for simple problems, such as normal means with and without known variance, difference between two proportions.
  • Introduction to BUGS software ( Bayesian analysis Using Gibbs Sampling)
  • Notes from course pack.
Clinical trials
[5] Feb 5, 2001
  • Design issues in clinical trials.
  • Decision making during the course of a clinical trial.
  • Interim analyses in clinical trials.
  • Reporting Bayesian analyses in clinical trials.
  • Notes from course pack.
  • D. Spiegelhalter, L. Freedman and M. Parmer. (1995). Bayesian approaches to randomized trials. Journal of the Royal Statistical Society, Series A 157:387-416.
  • M. Hughes, Reporting Bayesian analyses of clinical trials, (1993), Statistics In Medicine, 1651--1663
  • P. Fayers, D. Ashby, and M. Parmar, Tutorial in Biostatistics: Bayesian Data Monitoring in Clinical Trials, (1997), Statistics in Medicine 16, 1413-30.
Practical Implementation II
[6] Feb 12, 2001
  • Bayesian inference using the Gibbs sampler for linear and logistic regression.
  • Missing covariate data
  • G. Casella and E. George, Explaining the Gibbs Sampler, (1992), American Statistician, 167--174
  • Notes from course pack.
Special Topics
[7] Feb 26, 2001
  • Selected topics.