Monday, January 6, 2014: 2:00 PM-5:00 PM
Tanglin IV (The Regent Hotel)
Course Type: Half Day
Course Level: Intermediate

Format Requirements: The class will involve a mixture of lectures and a practical session supported by the faculty and followed by a class-level walk through of the exercise. Participants who attend with a laptop will be able to undertake the practical example within the class. The course provides an introduction to probabilistic modelling and VOI and, therefore, no experience of these techniques is required. However, participants will be expected to have an understanding of cost-effectiveness analysis as a pre-requisite, as such it is advertised as an intermediate level course. The practical exercise will be undertaken in Excel.

Background: Probabilistic modelling and Value of Information (VOI) analyses are increasingly being undertaken across the SMDM community (and beyond) to estimate the value of additional research. This course aims to provide an introduction to both probabilistic modelling and VOI, presenting the theory, general concepts and approach in an intuitive and non-technical manner for those who have heard the terms and want to know more. Participants will learn what types of questions VOI techniques can be used to address, which specific techniques are required to address which questions, and the role of probabilistic modelling in forming the foundation of a subsequent VOI. In addition, a class practical will provide participants with the opportunity to undertake a basic EVPI calculation using excel.

This course provides an introduction to the principles and practice probabilistic modelling and Value of Information analysis. It is aimed at those who have seen the presentations and heard the terminology and now want to interpret the figures, understand the implications and use the techniques themselves . The course will cover the concepts of EVPI, EVPPI, EVSI and ENBS, detailing when each approach can be used and for which questions.  The foundations of VOI analysis on an appropriate probabilistic representation of uncertainty will be emphasised. Case studies will be used to illustrate the techniques where appropriate and participants will get the opportunity to undertake an exercise to calculate the expected value of perfect information. 

The objectives of the course are:

  • to de-mystify the techniques of Value of Information analysis
  • for participants to be able to comprehend and explain the results of VOI analyses
  • for participants to understand the different levels of VOI analysis and which questions can be addressed by each of the different techniques
  • to emphasise the role of probabilistic modelling of uncertainty as a basis for VOI
  • for participants to have the opportunity to undertake an Expected Value of Perfect Information exercise in excel

14:00 – 14:05 - Introduction to course
14:05 – 14:50 - Lecture session: Undertaking probabilistic analysis
14:50 – 15:30 - Lecture session: VoI principles                                                    
15:30 – 16:00 - Coffee break
16:00 – 16:45 - Practical session:  Probabilistic analysis & EVPI                      
16:45 – 17:15 – Lecture session: VoI interpretation & understanding       
17:15 – 17:30 - Wrap-up & Conclusions

Course Director:
Andrew Briggs, DPhil
Course Faculty:
Olivia Wu