Description
Decisions that involve multiple attributes are commonly supported by decision analysis methods deploying a multi-attribute value/utility function, which captures decision-makers’ preferences on multiple decision attributes and their risk attitude. Often, there are uncertainties related to the evaluation of the decision alternatives' attribute-specific performances. In the context of single-attribute problems, these uncertainties cause the ex post realized performances of the selected alternatives to be systematically lower than the ex ante estimates, causing the decision-maker to experience post-decision disappointment. We study the impact of attribute-specific estimation uncertainties and nonlinear utility functions on post-decision surprises and the quality of decision making in a multiattribute setting. In particular, we present both analytical and simulation results to analyze the performance of several selection approaches by examining whether these approaches lead to the optimal choice that maximizes the expected multi-attribute utility and whether these choices correspond to the truly optimal choices that maximize the ex post true multi-attribute utility. Furthermore, we demonstrate the impacts of these selection approaches in a realistic example on choosing a contractor for a construction project.| Period | 1 Jul 2024 |
|---|---|
| Event title | European Conference on Operational Research |
| Event type | Conference |
| Conference number | 33 |
| Location | Copenhagen, DenmarkShow on map |
| Degree of Recognition | International |