My question refers to examining present day method for performing sensitivity analysis in MCDM
In my opinion, the way it is performed and the results it achieves are very poor, and almost useless
Sensitivity analysis as performed today only informs when a ranking changes, due to changes in a weight of a criterion that is arbitrarily chosen, since it is SUPPOSED, without any rationality ,that the criterion with the highest weight is the most important, which by the way it is easily demonstrable that said assumption is a fallacy
It also indicates the sensitivity of the selected alternative related to criteria, that COULD be very useful if the pertinent criterion were significant for that alternative. Since this criterion is chosen arbitrarily, there is no guarantee about the sensitivity of the solution found.
What we need is a system that provide useful and DOCUMENTED information, such as for instance how sensitive is the alternative selected to changes in exogenous variable over which the company don't have any control