Αn empirical study to measure the use and impact of an imprecise information approach in the decision support system Entscheidungsnavi

In a rational multi-criteria decision-making (MCDM) process based on multi-attribute utility theory (MAUT), the decision-maker defines quantitative parameters to identify the most promising alternative. However, it is often difficult to determine these parameters precisely. Therefore, the literature suggests using imprecise information, which allows the decision-maker to determine imprecise preference statements and parameters. This paper investigates the utilization and impact of an imprecise information approach regarding probabilities, utility functions, and objective weights in the decision support system Entscheidungsnavi . We examined 1,511 personal student decisions. First, we analyzed how often imprecise information was utilized. Second, we surveyed how helpful participants found the use of imprecise information. Third, we investigated the impact of imprecise information in the three aforementioned categories on the final ranking of alternatives. We assessed how robust the rank of the best alternative remained by using Monte Carlo simulations in which evenly distributed values were drawn from all imprecisely defined intervals. This demonstrates the robustness of our results across varying levels of input uncertainty. Most participants chose imprecise parameters in their decisions. Moreover, most participants found it helpful to use imprecise information in all categories, with imprecise objective weights being the least helpful. In about 83% of the decisions, the best-ranked alternative is robust against the imprecise intervals the decision-maker chooses. Overall, our results show that incorporating imprecision improves the practical usefulness of MCDM-Support-Systems. We recommend that a robust MAUT-based system should address all three types of imprecision, guide probability and utility elicitation, support realistic trade-offs, and visualize how input uncertainty affects ranking stability helping decision-makers form sound judgements.