Ram, Camelia
(2012)
Extending the combined use of scenarios and multi-criteria decision analysis for evaluating the robustness of strategic options.
PhD thesis, London School of Economics and Political Science.
Abstract
Deep uncertainty exists when there is disagreement on how to model inter-relationships between variables in the external/controllable and internal/controllable environment; how to specify probability distributions to represent threats; and/or how to value various consequences. The evaluation of strategic options under deep uncertainty involves structuring the decision problem, specifying options to address that problem, and assessing which options appear to consistently perform well by achieving desirable levels of performance across a range of futures. The integrated use of scenarios and Multi-Criteria Decision Analysis (MCDA) provides a framework for managing these issues, and is an area of growing interest. This thesis aims to explore such integrated use, suggesting a new method for combining MCDA and scenario planning, and to test such proposal through a multi-method research strategy involving case study, behavioural experiment and simulation. The proposal reflects the three key areas of confluence of scenarios and MCDA in the decision making process. The first is based on systematic generation of a larger scenario set, focused on extreme outcomes, for defining the boundaries of the decision problem. The second proposal is based on providing less scenario detail than the traditional narrative, in favour of explicitly considering how uncertainties affect positive and negative outcomes on key objectives. This backward logic seeks to better address the challenge of estimating the consequences of each option and the trade-offs involved. Finally, it is proposed that option selection be based on a concern for robustness through cost-equivalent regret. The empirical findings reflect that the key benefit of integration appears to be a mechanism to improve the efficiency of elicitation and the robustness of options. However, effective application of scenarios and MCDA requires awareness of the desired degree of accuracy required and risk attitude of decision makers.
Actions (login required)
|
Record administration - authorised staff only |