Appendix D:
Multi-Criteria Decision-Making Methodologies
This appendix lists some methods for multi-criteria decision-making that can be useful in project and portfolio management. These methods are employed mostly for selecting projects within a portfolio, as well as for making important project decisions. Each of these methods has its own strengths and weaknesses (see Linkov et al. 2006a).
Selecting methodologies and tools for multi-criteria decision-making should be part of the decision analysis process within your organization. It would be better to use this approach for many problems within a portfolio rather than for one particular problem. A number of off-the-shelf software tools can be used for the various methods. In particular, some methodologies are implemented as part of project portfolio management software.
Short Description |
References |
|
Analytic Hierarchy Process (AHP) |
• Develop a hierarchy that includes decision alternatives and criteria. • Perform pair-wise comparison to establish consistent priorities for different criteria. Input data for pairwise comparison is an expert judgment. • Calculate overall score for different alternatives and rank them according to score. |
Anderson et al. 2015; Saaty and Vargas 2014 |
Goal Programming |
A linear programming approach to multi-criteria decision problems whereby the objective function is designed to minimize deviation from goals |
Anderson, et al. 2015; Schniederjans 2012 |
Multi-Attribute Utility Theory |
1. Derive single-attribute functions for project parameters, such as project duration and cost. 2. Combine single-attribute utility functions to create multi-attribute utility function. 3. Perform consistency check to verify that multi-attribute utility function actually represents decision-maker’s preferences. |
Goodwin 2014; Keeney and Raiffa 1993 |
Simple Multi-Attribute Rating Technique (SMART) |
1. Construct value trees, which represent decision-making criteria. 2. Define value functions, which represent relationships between criteria (e.g., project cost vs. project value). 3. Determine weights for all criteria. Compute overall value (score) for each alternative. 4. Perform sensitivity analysis to determine how sensitive value is to the selected weights. |
Goodwin 2014 |