EURO 2025 Leeds
Abstract Submission

3090. Multinomial Decision Making (MNDM): A new Multi Criteria Decision Analysis based on random utility theory

Invited abstract in session MB-8: Preference Learning 1, stream Multiple Criteria Decision Aiding.

Monday, 10:30-12:00
Room: Clarendon SR 2.08

Authors (first author is the speaker)

1. David Palma
Management, Leeds University Business School
2. Richard Hodgett
Leeds University Business School, University of Leeds
3. Sajid Siraj
Leeds University Business School, University of Leeds
4. Romain Crastes dit Sourd
University of Leeds

Abstract

Common limitations in Multi Criteria Decision Analysis (MCDA) methods include (i) the need for direct definition of criteria weights, even though the literature on economics often favours indirect measures of preferences, e.g. in WSM, MAVT, TOPSIS and SMART; (ii) the often large number of pairwise comparisons required by methods like AHP or MACBETH; (iii) the need to normalise the value of criteria in methods like WSM, TOPSIS and SMART, introducing an additional source of variability when handling negative or minimising criteria; and (iv) the lack of flexibility dealing with uncertainty and interactions of criteria.

Multinomial Decision Making (MNDM) is a new method based on discrete choice modelling techniques and random utility theory. It requires the user to make a series of hypothetical simple choices (allowing for equal preference) among synthetic alternatives defined from subsets of criteria. From these choices, the weights of criteria are derived indirectly and a utility function constructed. The number of choices grows linearly with the number of criteria, avoiding overwhelming the user. Furthermore, it does not require normalising criteria values, allowing for the measurement of explicit trade-offs among criteria. A flexible definition of the utility function allows for interactions and non-linear effects of criteria (e.g. ideal point solutions). An explanation of the method and a software implementation will be presented at the conference.

Keywords

Status: accepted

Back to the list of papers