Paper Type
Complete
Paper Number
PACIS2026-1968
Description
In information system and cloud vendor selection, decision support systems (DSS) influence how confidently managers act on recommendations. If uncertain evaluations are represented as symmetric intervals, implicitly assuming the expected value lies at the midpoint, the resulting ranking may convey a false sense of decisional precision. This paper designs a simulation-based DSS artifact extending Simple Additive Weighting (SAW) with Asymmetric Interval Numbers (AINs), representing each criterion score as a lower bound, expected value, and upper bound. Monte Carlo simulation propagates directional uncertainty through the aggregation model. The artifact is demonstrated in a cloud vendor selection case against deterministic and symmetric-interval SAW. Results show that symmetric modeling can distort top-rank probabilities. A sensitivity analysis identifies an indicative threshold: when the asymmetry coefficient exceeds 0.3 for highly weighted criteria, symmetric and asymmetric recommendations begin to diverge qualitatively. The study contributes a transparent DSS artifact and guidance for representing uncertainty in procurement-oriented DSS.
Recommended Citation
Sałabun, Wojciech; Kizielewicz, Bartłomiej; Bączkiewicz, Aleksandra; and Wątróbski, Jarosław, "When Symmetry Misleads in DSS: Modeling Directional Uncertainty with AIN-Based SAW for Information Systems Selection" (2026). PACIS 2026 Proceedings. 13.
https://aisel.aisnet.org/pacis2026/practioner/practioner/13
When Symmetry Misleads in DSS: Modeling Directional Uncertainty with AIN-Based SAW for Information Systems Selection
In information system and cloud vendor selection, decision support systems (DSS) influence how confidently managers act on recommendations. If uncertain evaluations are represented as symmetric intervals, implicitly assuming the expected value lies at the midpoint, the resulting ranking may convey a false sense of decisional precision. This paper designs a simulation-based DSS artifact extending Simple Additive Weighting (SAW) with Asymmetric Interval Numbers (AINs), representing each criterion score as a lower bound, expected value, and upper bound. Monte Carlo simulation propagates directional uncertainty through the aggregation model. The artifact is demonstrated in a cloud vendor selection case against deterministic and symmetric-interval SAW. Results show that symmetric modeling can distort top-rank probabilities. A sensitivity analysis identifies an indicative threshold: when the asymmetry coefficient exceeds 0.3 for highly weighted criteria, symmetric and asymmetric recommendations begin to diverge qualitatively. The study contributes a transparent DSS artifact and guidance for representing uncertainty in procurement-oriented DSS.
Comments
16-Practitioner