SCSlogo

 

 

 

 

 

 

 

“Why a Bayesian Be?”

by Andrew R Willan, PhD
Senior Scientist Emeritus, Sick Kids Research Institute
Professor, Dalla Lana School of Public Health, University of Toronto

Date: Friday, April 8, 2016
Time: 1:00pm-3:00pm
Location: York Lanes 283B (enter through York Lanes 283)

Abstract:
Bayesian approaches to the design, analysis and interpretation of clinical studies have important advantages. They provide a transparent
means of combining prior evidence with new data. They also allow direct statements of inference regarding model parameters that are simple,
relevant and intuitive. Further, Bayesian methods provide the basis for decision analysis addressing questions such as: is the current evidence
in favour of a new health intervention sufficient for adoption, and if not, what is the optimal design for future research? Examples will be
used to illustrate the importance of these advantages, illustrating how Bayesian methods provide statements of inference which are superior to
those provided by frequentist methods. Lastly, a brief introduction to value of information methods (decision analysis) will be given.

Sponsored by the Statistical Consulting Service and the Institute for Social Research.