Speaker
Description
Many important decisions face a basic dilemma: act now, while the facts are still uncertain, or wait for better information at the cost of delay. Standard Multi-Criteria Decision Analysis (MCDA) does not address this dilemma directly, since it treats the performance of each option as a fixed number evaluated at one moment in time, and says nothing about when that moment should be. This paper proposes a new framework, Time-Dynamic MCDA (TD-MCDA), that treats the decision moment itself as something to be chosen. It captures two forces that pull in opposite directions. As time passes, decision-makers typically learn more and become more confident about how each alternative will actually perform, so uncertainty falls. But every extra period of inaction can bring its own cost, whether in the form of missed benefits, ongoing harm, or accumulating damage, so a penalty for delay rises. We formalise both forces mathematically and combine them into a single trade-off that identifies the best time to decide: the point beyond which any further delay would cost more than the extra certainty it buys. The framework works with any way of expressing uncertainty, from simple ranges to full probability distributions to fuzzy numbers, so it can be adapted to many types of problems. We illustrate it with a practical example, choosing where to site a new solar power plant, showing how the model weighs the environmental cost of delay against the value of waiting for clearer forecasts, and recommends both a specific point in time to decide and which site to choose. We close by outlining several extensions, including versions that let the decision be revisited as new evidence arrives.
| Keywords | Decision support systems; MCDA; Decisions under uncertainty |
|---|---|
| Participation in the SNSF Early-Stage Researcher Support and Award Scheme | I am submitting a discussion paper; I understand that discussion papers are not eligible for the scheme. |