Article type
Year
Abstract
Objectives: The purpose of this workshop is to introduce Bayesian meta-analysis (BMA), convey its scope and importance, followed by a hands-on training in WinBUGS. At the end of the workshop, participants will be sensitized to the concepts of BMA and receive practical tips on performing BMA.
Description: The Bayesian methods are gaining prominence in the domain of evidence synthesis due to increased awareness and advent of computational software packages. However, their acceptance has not been as wide as the frequentist meta-analysis as they are perceived to involve complicated computational procedures. We are proposing an introductory workshop on BMA consisting of three parts. Part 1 will focus on the basic concepts of BMA: prior distribution, likelihood function, posterior distribution, fixed- and random-effects model (10 minutes). Part 2 sensitizes on choosing prior distributions (10 minutes), and Part 3 involves hands-on training on the use of WinBUGS software with simple practical examples focussing on binary and continuous outcomes, which is followed by interpretation (60 minutes). The final 10 minutes will be for interaction/discussion. Participants will receive a handout consisting of detailed description of Bayesian meta-analysis procedure and WinBUGS codes. Participants should bring or share a laptop for hands-on exercises. WinBUGS should be installed prior to the workshop (www.mrc-bsu.cam.ac.uk/software/bugs/the-bugs-project-winbugs).
Description: The Bayesian methods are gaining prominence in the domain of evidence synthesis due to increased awareness and advent of computational software packages. However, their acceptance has not been as wide as the frequentist meta-analysis as they are perceived to involve complicated computational procedures. We are proposing an introductory workshop on BMA consisting of three parts. Part 1 will focus on the basic concepts of BMA: prior distribution, likelihood function, posterior distribution, fixed- and random-effects model (10 minutes). Part 2 sensitizes on choosing prior distributions (10 minutes), and Part 3 involves hands-on training on the use of WinBUGS software with simple practical examples focussing on binary and continuous outcomes, which is followed by interpretation (60 minutes). The final 10 minutes will be for interaction/discussion. Participants will receive a handout consisting of detailed description of Bayesian meta-analysis procedure and WinBUGS codes. Participants should bring or share a laptop for hands-on exercises. WinBUGS should be installed prior to the workshop (www.mrc-bsu.cam.ac.uk/software/bugs/the-bugs-project-winbugs).