2. Multi-Criteria Analysis

Multi-Criteria Analysis aims at integrating the different dimensions of reality into a single analysis framework, in order to give an integral vision of the situation or problem considered. 

At a first level, we can visualise it as an impact matrix or evaluation matrix where the lines, for instance, represent the different criterias against which the alternatives or scenarios in columns are evaluated.

A scenario may perform better (be better evaluated) for criterion 1, but perform less well for a criterion 2.

Evaluation matrix
  Alternative A Alternative B Alternative C Alternative D
Criteria 1 Evaluation for 1A Evaluation for 1B Evaluation for 1C Evaluation for 1D
Criteria 2 Evaluation for 2A Evaluation for 2B Evaluation for 2C Evaluation for 2D
Criteria 3 Evaluation for 3A Evaluation for 3B Evaluation for 3C Evaluation for 3D
Criteria 4 Evaluation for 4A Evaluation for 4B Evaluation for 4C Evaluation for 4D
Criteria 5 Evaluation for 5A Evaluation for 5B Evaluation for 5C Evaluation for 5D

Evaluation of each cell may be done through different types of metrics. 

G. Munda set up in 1995 a five step process for multi-criteria analysis:

  1. Definition and structuring of the problem to investigate
  2. Definition of a set of criteria for evaluation
  3. Choice between discrete and continuous methods
  4. Identification of preferences of the decision-maker
  5. Choice of the aggregation methods for the criteria

Discrete and continuous methods (step 3) depend on the finite or infinite number of alternatives.

Agregation methods in MCA (step 5) are diverse. Several methods can be mentionned:

 

The objectives of a multicriteria analysis are diverse, for example they can be one or several of the following:

- To identify a preferred alternative

- To allocate each alternative to a pre-defined category, according to the objectives of the process, for example: accept, reject, reserve list.

- To order the alternatives, following a complete or a partial ordering, for example:

            A > C > B > D  or A,C > B > D

-          To describe the relevant alternatives and their consequences

Formalized methods lead the analyst of multi-criteria issues to the objective sought.

For an example, you can refer to the MCA of the Yasuni ITT initiative (its limitations are in the premises chosen, not in the method).