How do you elicit and aggregate preferences from multiple stakeholders in MCDA?

1 Define the problem and the criteria

The first step in any MCDA process is to define the problem and the criteria that will be used to evaluate the alternatives. This step requires a clear understanding of the decision context, the objectives, and the constraints. It also requires a consultation with the relevant stakeholders, who may have different perspectives and expectations. A good way to define the problem and the criteria is to use a structured framework, such as SMART (Simple Multi-Attribute Rating Technique), which helps to identify, measure, and weight the criteria.

2 Identify the stakeholders and their roles

The next step is to identify the stakeholders and their roles in the decision process. Stakeholders are the individuals or groups who are affected by or have an influence on the decision. They may have different levels of involvement, interest, and power. A stakeholder analysis can help to map out the stakeholders and their roles, such as decision makers, advisors, beneficiaries, or opponents. A stakeholder analysis can also help to determine the appropriate methods and formats for eliciting and aggregating their preferences.

3 Elicit the preferences of the stakeholders

The third step is to elicit the preferences of the stakeholders, which means to collect their opinions, values, and judgments on the criteria and the alternatives. There are various methods for eliciting preferences, such as surveys, interviews, focus groups, workshops, or interactive tools. The choice of the method depends on the number, diversity, and availability of the stakeholders, as well as the complexity and uncertainty of the problem. The main goal of eliciting preferences is to capture the trade-offs, weights, and scores that reflect the stakeholders' views and priorities.

4 Aggregate the preferences of the stakeholders

The fourth step is to aggregate the preferences of the stakeholders, which means to combine them into a single or multiple preference models that can be used to rank or compare the alternatives. There are different approaches for aggregating preferences, such as additive, multiplicative, outranking, or consensus models. The choice of the approach depends on the type, scale, and distribution of the preferences, as well as the degree of agreement or conflict among the stakeholders. The main challenge of aggregating preferences is to balance the representation, fairness, and robustness of the preference models.

5 Analyze the results and sensitivity

The fifth step is to analyze the results and sensitivity of the preference models, which means to examine the outcomes, rankings, or recommendations that are derived from the preferences of the stakeholders. This step involves checking the validity, consistency, and reliability of the preference models, as well as exploring how they change under different scenarios, assumptions, or uncertainties. A sensitivity analysis can help to identify the key factors, criteria, or stakeholders that influence the results, as well as to test the robustness and stability of the preference models.

6 Communicate and validate the results

The final step is to communicate and validate the results of the preference models, which means to present and explain them to the stakeholders and other audiences. This step requires a clear and transparent communication strategy, which may include reports, presentations, dashboards, or visualizations. The communication strategy should also include feedback mechanisms, which allow the stakeholders to review, comment, or revise the results. The feedback mechanisms can help to validate the results, as well as to improve the trust, acceptance, and legitimacy of the preference models.