Sensitivity analysis of a simple additive weight method

Abstract

The authors depict the optimal criterion linked to sensitivity analysis. The aim is however to show case to multi-attribute utility analysis. Therefore, the authors go out from Simple Additive Method (SAW), but come to normalisation by the use of ratios where test matrices are formed. It is concluded that only significant attributes relative to the other attributes have an impact on the results. The proposed model for determining method sensitivity to changes of separate parameters enables us to increase the reliability of the applied methods.