Bayesian Statistics [EL003]

Detailed information about this course:

Description

There is growing acknowledgement of the value of Bayesian methods for complex models in biostatistics and epidemiology, in dealing with issues such as multiplicity, measurement error, spatial associations and hierarchical structure. This course will introduce the essentials of Bayesian ideas, emphasizing practical application using exact and simulation-based software. Examples will include the use of Bayesian methods in clinical trials, institutional comparisons, smoothing of disease rates, and frailty models.

Objectives

• Introduce Bayesian concepts and contrast them with the frequentist approach;
• Understand the background and the importance of computer intensive methods in Bayesian statistical analyses, such as the Markov Chain Monte Carlo (MCMC) techniques: Gibbs and Metropolis-Hastings sampling;
• Be able to work with some Bayesian software, such as WinBUGS etc.;
• Understand the applications and developments in the epidemiological literature involving Bayesian approaches.Keywords: Bayesian analysis, Gibbs sampling, MCMC.
  

Participant profile

Epidemiologists and clinical researchers.

Assessment

Assignment(s), Attendance