COST

Decisions are inevitably based on preferences (values) and beliefs (estimates) of the stakeholders taken into account in a decision process. Although COST Action IS1304 focuses mainly on eliciting beliefs about uncertain outcomes, the Elicitation of stakeholder preferences theme within WG1 focuses on preference elicitation. Preference elicitation is a socio-technical process aiming at setting the value of preference-related parameters in decision aiding models. Such parameters are typically subjective involving some value judgement. For instance, a preference-related parameter can be the weight of a criterion in a multi-criteria evaluation model, a risk-aversion parameter in a utility theory framework, or a target to be set.

The following arguments justify the relevance of this theme for this COST action:

• Expert judgment about uncertain outcomes occurs within a larger decision making context. If an estimate is sought, this is a consequence of stakeholder preferences determining that a given decision objective is to be considered, and determining that it should be assessed through that estimate. After an estimate is elicited, the contribution (“weight”) that this estimate will have within the larger decision making context will again be determined by elicited stakeholder preferences.
• Preferences can be influenced by beliefs, especially if the consequences of the decision will span a long period of time. Trade-offs between economic and environmental consequences, for instance, are likely to be affected by expectations about the future concerning these dimensions.
• Methods and processes designed for eliciting expert judgement about uncertainties may be applicable (possibly with adaptations) to stakeholder preference elicitation, and vice-versa.

In what concerns the theme Elicitation of stakeholder preferences, WG1 researchers aim at improving preference elicitation processes through new scientific development, training, and the spread of good practice. Preference elicitation processes are typically challenging even when the preferences of a single individual are considered: such preferences are not perfectly constructed in the individual’s mind, they are subject to elicitation biases, and need to be adequately translated to the language (usually mathematical) of the decision aiding model. When preferences concern multiple stakeholders, the process is even more complex, because preferences can be diverse, there may be hidden agendas, and group dynamics must be managed. When the preferences concern stakeholders with whom it is not possible to interact (e.g., Europe’s population, or future generations), then one way forward is to rely on expert judgement – precisely the main topic of this COST Action.

A (provisional) list of topics of interest within the Elicitation of stakeholder preferences theme is:

• Elicitation methods: What approaches work well? How to minimize biases?
• Robustness Analysis to elicitation methods: How to address the above issues by promoting solutions, which are acceptable under a wide set of hypotheses, assumptions and estimates?
• Indirect elicitation approaches: inferring parameter values from examples of decisions.
• Elicitation of preferences in groups in facilitated workshops: What approaches work well? How to minimize phenomena such as inhibition or groupthink?
• Elicitation of preferences in groups dispersed in time and space:  What approaches work well? How can technology (for instance Internet-based systems) help?
• Methods to elicit the preferences of a population: how can these be sampled and estimated?
• Methods to elicit future preferences: how can expert judgement be assessed and used?
• Forecasting the preferences of others (adversaries) in security contexts: How to leverage expert judgement?
• How can methods developed for eliciting estimates be adapted for the elicitation of preferences or vice-versa? 

Argyris, Nikos

Aspinall, Willy

Cooke, Roger

Delias, Pavlos

Dias, Luis

French, Simon

Grigoroudis, Evangelos

Kasemets, Aare

Liesio, Juuso

Mackay, Cameron

Matsatsinis, Nikolaos

Mikulskiene, Birute

Neslo, Rabin

Qazi, Abroon Jamal

Rios Insua, David

Salo, Ahti

Smith, Jim

Srdjevic, Zorica

Walker, Jamie

Zdziarski ,Michał

Rendtorff, Jacob Dahl