The lazy assumption when faced with several different criteria is to multiply each option’s score against a criteria by the weight given to that criteria and add the results (this is also mathematically invalid unless all are measured at the ratio scale).
Technically, this is an additive Multi-Attribute Utility function. It is one form of a group of functions termed Compensatory; thus, if one option performs exceptionally badly against one criteria but well against all others, it will still score reasonably overall. This can lead to very poor decisions! Non-compensatory forms of this kind of analysis set minimum pass-fail standards for specific criteria, to avoid such decisions.
At the beginning of the analysis, the actual form of the relationship between the quantities of some good and the ‘desirabilities’ of those quantities, the utility function for each good or attribute, will not be known. More particularly, the way in which the individual combines the utility functions for the many different goods or attributes, the Multi-Attribute Utility function, is not known. Consequently it will not be known whether they all share the same functional form. So, the two questions are:
• Does the functional form make a difference to the identification of the best option? and,
• Consequently, does it matter if stakeholders disagree?