This course aims to introduce students to the field of decision analysis. At the end of the course students should be able to develop solutions to basic decision problems involving uncertainty, using specialised, free-ware software. An intuitive exposition into Bayes’ theorem is given through a number of simple-to-follow problems involving information update in light of new evidence, to show the importance of this theorem in making decisions about risk. Students learn about subjective probabilities and various formal methods of expert opinion elicitation. Two specific decision tools - Bayesian networks and Multi-criteria Decision Analysis (MCDA), are examined in detail.

Bayesian networks, which display graphically the interrelationships between variables, allow for their dependencies to be accounted for in a manner reflecting their causal relationships. This permits the evaluation of predicted consequences associated with different interventions, before these are actually implemented. MCDA allows decisions to be made when the decision-maker is faced with multiple and conflicting objectives where tradeoffs must be made, which is most often the case in the real world. Netica and SuperDecisions are free-ware software implementing Bayesian networks and MCDA, respectively, which students will use during the course to solve decision problems from a range of application fields.

Finally the course will discuss current issues in behavioural risk management and the value of first-line employees’ experience, such as their ability to identify ‘early warnings’ about risk, and what an organisation can do to encourage such employees to speak up without the fear of retribution. The aim is to demonstrate the benefits of front-line employees taking charge of local risk management issues and the importance role risk culture plays in allowing this to occur.