A collaborative group decision-support system: the survey based multi-actor multi-criteria analysis (MAMCA) software

Among various stakeholder-based multi-criteria group decision-making (MCGDM) frameworks, Multi-Actor Multi-Criteria Analysis (MAMCA) stands out for its emphasis on exploring preferences and conflicts among stakeholder groups. MAMCA ensures comprehensive stake-holder assessments, integrating diverse perspectives into the decision-making process. However, existing MAMCA software faces limitations in accommodating large participant bases, a challenge amplified by the increasing demand for online stakeholder workshops. This study extends the MAMCA software by incorporating survey-based evaluation methods, evolving it into a survey-based decision-support system (DSS). This enhanced DSS facilitates output-level aggregation within stakeholder groups, enabling detailed collection and aggregation of individual preferences. The survey-based DSS includes new developed survey-based techniques such as the Simos method and Simple Multi Attribute Rating Technique (SMART), which streamline the weight elicitation and alternative appraisal processes. These methods ensure an intuitive and accessible evaluation process, suitable for both online and offline workshops. A case study evaluating Swiss energy policies demonstrates the effectiveness of the survey-based DSS, highlighting its ability to support comprehensive stakeholder assessments and facilitate understanding across diverse groups. This approach improves stakeholder engagement and promotes transparency throughout the decision-making process, ensuring inclusive decision outcomes.


Introduction
Decision-making within social contexts often grapples with the challenge of integrating multifaceted criteria and diverse stakeholder perspectives (Soltani et al., 2015).Multi-Actor Multi-Criteria Analysis (MAMCA) stands out as a group decision-making (GDM) framework designed to address this complexity by actively involving various stakeholder groups in the decision-making process (Macharis et al., 2009).MAMCA facilitates a structured evaluation of alternatives, where different stakeholder groups can apply unique sets of criteria reflecting their distinct priorities and concerns.This approach not only enhances the relevance and acceptability of decisions but also fosters collaborative problem-solving among stakeholders with potentially conflicting interests (Macharis et al., 2012).
Recognizing the need to support interactions with stakeholders, specialized MAMCA software has been developed (Huang et al., 2020).This software aims to facilitate the decision-making process by providing a platform for interaction and evaluation.Nevertheless, the MAMCA software has encountered limitations, particularly in accommodating large-scale participatory settings and effectively engaging stakeholders.These issues have been exacerbated by the growing trend towards digital and remote collaboration, driven by wider changes in communication practices (Cutler et al., 2021).
This paper introduces an innovative extension of the MAMCA software, incorporating survey-based evaluation methods to transform it into a more scalable and user-friendly decision-support system (DSS) (Razmak & Aouni, 2015;Zhou et al., 2004).This enhancement is aimed at overcoming the barriers to broad stakeholder engagement by facilitating easier and more inclusive participation.The effectiveness of this enhanced system is demonstrated through a case study on Swiss energy policy, illustrating how the revised tool aids in achieving a deeper understanding and alignment among stakeholders, thereby leading to more informed and sustainable decision-making outcomes.
The organization of this paper is structured as follows: Section 2 presents the literature review, which explores the important role of stakeholder involvement in decision-making and the positioning of MAMCA within the framework of stakeholder-based group decision-making.In Section 3, we introduce the MAMCA methodology and its software, detailing the workflow and functional aspects of MAMCA.In Section 4, we outline the framework for implementing MAMCA through a survey-based approach.Section 5 features a real-life case to demonstrate the feasibility and visualisation capabilities of our DSS based on MAMCA software.Finally, Section 7 offers concluding remarks.

Literature review
In addressing a social decision-making problem such as urban planning (Ali-Toudert et al., 2020) and policymaking (Wallace et al., 2020), the complexity is multifaceted.For example, in energy policymaking, this complexity arises from the need to integrate various dimensions-resilience, security, sustainability (Cinelli et al., 2022).Stakeholder engagement is crucial, as policies must address the concerns and interests of diverse groups ranging from local communities to multinational corporations (Ortiz & Leal, 2020).Moreover, aligning short-term actions with long-term goals adds to the challenge, requiring adaptive strategies that can evolve with changing circumstances and insights (Kanellakis et al., 2013).
During the decision-making process, Decisionmakers (DMs) are tasked with the challenge of comprehensively evaluating various dimensions of the issue (Almeida et al., 2020;Howdon et al., 2022).When evaluating possible alternatives in social decision-making problems, one alternative is unlikely to dominate the others, ie, a choice that outperforms the other options on every criterion (Uzun et al., 2021).Thus, decision-makers frequently grapple with the challenge of reconciling trade-offs among conflicting criteria.This dilemma is recognised as a multiple-criteria decision-making (MCDM) problem within the operational research literature (Zionts, 1979).MCDM methods have been developed to structure and systematically address these multifaceted problems.The objective of multi-criteria approaches is to guide decision-makers in selecting from a range of alternatives, each evaluated based on a set of criteria.This approach supports decision-making by offering different functionalities: it can rank alternatives from most to least favourable, sort them into distinct categories based on predefined criteria, facilitate the selection of the most suitable alternative, or cluster similar alternatives together (Belton et al., 2002).MCDM offers benefits compared to methods like cost-benefit analysis (CBA), which typically focuses on generating a single economic value (Damart & Roy, 2009).By accommodating a broader range of considerations, MCDM facilitates a more thorough and nuanced evaluation, promoting a deeper insight into the available alternatives and their respective consequences (Annema et al., 2015).
Another complexity of social decision-making problems lies in the involvement of different stakeholders (Freeman, 2010).These stakeholders include various parties and individuals who are conscious of the impact these decisions will have on them.When individuals and organizations either influence or are impacted by a decision, they constitute distinct stakeholder groups associated with the problem (Freeman et al., 2010).Stakeholders in social decision-making encompass a broad range of parties including government entities, businesses and industries, NGOs, the public, etc (Gregory et al., 2020;Pluchinotta et al., 2022).It's crucial to acknowledge and integrate the varied viewpoints and priorities of stakeholders for several reasons: 1. Enhanced knowledge and perspectives: Stakeholders contribute diverse perspectives that can lead to a more comprehensive understanding of issues.They bring unique insights and values to the table (Lukasiewicz & Baldwin, 2017).2. Acceptance and legitimacy: The involvement of stakeholders builds trust and acceptance, enhancing the legitimacy of decisions (Marleau Donais et al., 2021;Pluchinotta et al., 2022).3. Improved quality of decisions: Diverse inputs can lead to innovative solutions and avoidance of potential conflicts or errors (De Gooyert et al., 2017;Gregory et al., 2020).4. Viability of outcomes: When stakeholders are involved, they are more likely to support and sustain the decision outcomes.Their support can critically influence the successful implementation of decisions (Thabrew et al., 2009).
Stakeholders are pivotal in ensuring that decision-making processes are aligned with the interests and needs of those affected.Acknowledging their role not only improves the implementation of decisions but also addresses the ethical considerations of fairness and inclusivity (De Colle, 2005;Webler & Tuler, 2021).Therefore, correctly eliciting stakeholders' preferences, efficiently engaging them, and obtaining their information becomes important (Cairns et al., 2016;Zheng & Lienert, 2018).
In a multi-criteria problem, involving stakeholders transforms the issue into a group decision-making (GDM) scenario, more specifically termed as a stakeholder-based multi-criteria group decision-making (MCGDM) problem (Chakhar & Saad, 2014;Song & Hu, 2019).Stakeholder-based MCGDM has become increasingly popular in addressing social decisionmaking problems (Hwang & Lin, 2012).Within these frameworks, facilitated modelling is a critical approach for supporting complex decision-making processes (Franco & Montibeller, 2010).This method involves active stakeholder participation, ensuring that their perspectives are incorporated into the decisionmaking framework (McCartt & Rohrbaugh, 1989).Practitioners guide the group through structured problem-solving and decision-making processes, using tools such as problem structuring methods, system dynamics, and decision analysis (Schilling et al., 2007).These methods enhance collaborative efforts by promoting full participation, mutual understanding, and inclusive solutions, making them particularly suitable for complex strategic decisions (McCartt & Rohrbaugh, 1995).MCDM methods are integral to facilitated group decision-making, providing structured frameworks for eliciting, weighting, and aggregating stakeholder preferences, and supporting both qualitative and quantitative aspects of decision-making (Bana e Costa et al., 2002;Bana e Costa et al., 2014;Parnell et al., 2013).These methods ensure that diverse stakeholder inputs are systematically considered, leading to more robust and transparent decision outcomes.In the following subsection, we discuss various MCGDM frameworks at a high level, highlighting their structural and procedural similarities and differences.

Stakeholder-based MCGDM frameworks -Their homogeneity and heterogeneity
As a group decision-making extension of MCDM, various stakeholder-based MCGDM frameworks have been developed to involve stakeholders and aggregate their preferences.It is essential to examine both the homogeneity and heterogeneity across different frameworks.This analysis will not only highlight the common functional features and underlying principles they share but will also identify their distinct characteristics and implementation variations.Such a comparison is crucial for understanding how each framework accommodates different stakeholder needs, adapts to various decision-making contexts, and handles the complexity of group dynamics.We have identified the stakeholder-based MCGDM frameworks that are well-established in the literature.These frameworks are detailed in methodological papers that have accumulated over 100 citations in the Scopus database (Guz & Rushchitsky, 2009): � Multi-Actor Multi-Criteria Analysis (MAMCA) (Macharis et al., 2009): Engages multiple stakeholder groups, allowing them to use their own sets of criteria to evaluate alternatives and illustrate inter-group conflicts.� Multi-Criteria Mapping (Stirling & Mayer, 2001): Allows participants to explore the broader implications of various policy options by assessing them against a diverse set of criteria explicitly chosen by the stakeholders.� Social multi-criteria evaluation (SMCE) (Munda, 2004): Integrates social and ethical considerations by combining quantitative and qualitative data to address complex policy issues.� Multi-criteria evaluation (MCE) (Stagl, 2006): Applies multiple criteria to evaluate and compare options in a structured decision-making process, highlighting the trade-offs between competing objectives.� Deliberative Multicriteria Evaluation (Proctor & Drechsler, 2006): Combines quantitative assessment with deliberative processes to facilitate dialogue among stakeholders, enriching the decision-making process.� Decision analysis interview (Marttunen & H€ am€ al€ ainen, 1995): Involves structured individual or group interviews designed to elicit preferences and judgments about decision options and criteria.� Participatory MCDA (Nordstr€ om et al., 2010): Enhances the engagement of stakeholders by allowing them to actively participate in defining the criteria and evaluating the alternatives, fostering transparency and inclusiveness.
There are also other stakeholder-based MCGDM frameworks such as Eskandari et al. (2012); Roussat et al. (2009); Scott et al. (2015); Sheppard and Meitner (2005).However, these frameworks are more generalized and less well-defined, making them less relevant to our focused discussion.First of all, all the frameworks share a similar evaluation process.Kabak and Ervural (2017) discuss the process of the MCGDM framework, and for the stakeholder-based MCGDM process the process is similar: 1. Structuring and construction stage: This initial stage sets the groundwork for the MAGDM process.a. Identify the decision goal: Clearly define the purpose and objectives of the decisionmaking process.b.Form a committee of stakeholders: Assemble a group of experts or stakeholders who will participate in the decision process.c.Determine alternatives: List all possible options or actions that could be taken in response to the decision goal.d.Determine criteria: Establish the criteria that will be used to evaluate the alternatives.This could be based on the group consensus or dictated by the problem's nature.Although sharing a similar structure, the stakeholder-based MCGDM have different types of aggregation levels, and it impacts the outcomes of the decision-making process (Te Boveldt et al., 2021).From input-level aggregation to output-level aggregation, it influences how the preferences of stakeholders are expressed and incorporated into the final decision.Input-level aggregation prioritizes reaching consensus on various decision elements such as criteria, preferences, and criteria weights to generate a unified ranking of alternatives, similar to traditional MCDM processes.SMCE, MCE, and Deliberative Multicriteria Evaluation all belong to the input-level aggregation framework.Conversely, output-level aggregation segments stakeholders into distinct clusters based on variables such as socioeconomic background or geographic location.These clusters then undergo separate preference elicitation processes, aggregating global preferences.Decision Analysis Interview, Participatory Multicriteria Evaluation, and Participatory MCDA all belong to the output-level aggregation framework.Also, MAMCA and Multi-Criteria Mapping provide flexibility regarding where and how information is aggregated.Each stakeholder group submits individual inputs.However, within these groups, stakeholders can determine their collective preferences informally, often reaching a consensus through discussion.Additionally, MAMCA is distinct in that it typically does not aggregate collective preferences among stakeholder groups (Macharis et al., 2009).
Building on this foundational understanding, it is evident that the level of aggregation significantly shapes how stakeholder preferences are expressed and utilized within the decision-making process.The aggregation framework chosen not only determines the directness of stakeholder influence but also affects the levels of details that can be extracted and acted upon.In input-level frameworks where consensus on decision elements is critical, there is a concerted effort to harmonize differing stakeholder views into a cohesive decision strategy.This often involves intensive negotiations and deliberations to align diverse interests and priorities, thus ensuring that the decision reflects a collective agreement.
On the other hand, output-level frameworks acknowledge and preserve the distinctiveness of stakeholder groups by allowing them to express their preferences independently.This method can be particularly advantageous in complex decision environments where stakeholders represent varied and potentially conflicting interests.By segmenting the stakeholders, the framework can capture a broader spectrum of perspectives and facilitate a more equitable decision-making process that respects each group's unique contexts.
We can see that the different aggregation-level present serve for different scenarios.As already discussed, some frameworks provide flexibility of the aggregation.Among them, MAMCA is a more flexible framework, blending elements of both input and output-level aggregations.In MAMCA, the framework can be considered as an output-level aggregation framework, where each stakeholder group operates with its own criteria and evaluation processes.Final preferences for alternatives are aggregated, such as the case of assessing stakeholder support for different biofuel options in Belgium (Turcksin et al., 2011).Furthermore, MAMCA introduces the concept of a "group of stakeholders".Within such groups, several representatives may be involved.These intragroup representatives engage in input-level aggregation, whereas intergroup relationships follow an output-level aggregation, as demonstrated in the study of stakeholder preferences for low-carbon transport policies in China (Sun et al., 2015).More commonly, MAMCA encourages inputlevel aggregation within groups but does not aggregate preferences across different groups, thus allowing space for extensive roundtable discussions.
Thanks to its flexibility, MAMCA is frequently employed in the fields of transportation, mobility (Blad et al., 2022;Hamadneh et al., 2022), energy (Heuninckx et al., 2022;Lode et al., 2022), and environment (Almeida, 2019).MAMCA offers a more dynamic interaction between stakeholders and the decision process.By allowing stakeholders to input their preferences both individually and as part of a group, these models cater to the need for both personalized expression and group consensus.This flexibility can lead to richer data collection and a more nuanced understanding of stakeholder needs, enhancing the overall quality of the decision outcomes.Another notable aspect of MAMCA is the categorization of DMs and policymakers as stakeholder groups within the problem framework.Similarly, in social decision-making contexts, citizens are also recognized as stakeholder groups.This classification allows public preferences to be directly expressed, enabling citizens to actively participate in discussions (Brusselaers et al., 2021;Keseru et al., 2021).
However, due to MAMCA's high flexibility, there is a significant need for guidance from practitioners.Properly steering stakeholders through the decisionmaking process becomes particularly crucial in scenarios that combine input-level and output-level aggregations.There is a need for a more detailed exploration of the MAMCA structure, including its operational nuances, to ensure that all participant inputs are effectively integrated and managed throughout the decision process.

MAMCA and its software
As previously mentioned, MAMCA features a highly flexible structure for decision-making, supporting various types of preference aggregation.Therefore, practitioner support is essential at different stages of the MAMCA process to ensure clarity, inclusivity, and effectiveness.It is important to go through the MAMCA process to understand the need for practitioner support and the dynamic involvement of stakeholders.
Figure 1 illustrates the workflow of MAMCA, which aligns with the generic MCGDM framework process aforementioned.While MAMCA follows a similar process to the generic MCGDM framework, there are several key differences.One significant distinction is the concept of "stakeholder groups," which allows for the identification of multiple representatives within each group.These representatives perform input-level preference aggregation.Once the preferences of stakeholder groups are assessed, MAMCA can either proceed with output-level aggregation, if necessary, or directly involve stakeholders in a roundtable discussion to reach consensual decisions.
As shown in Figure 1.Practitioners play a critical role throughout this process.They help articulate a clear decision goal, identify and categorise stakeholders, and ensure fair representation.Practitioners also assist in brainstorming feasible alternatives, selecting relevant criteria, and facilitating systematic weight elicitation and alternative appraisals.During the preference aggregation and consensus-building phases, practitioners mediate discussions, reconcile conflicts, and guide stakeholders towards a consensus.Finally, they document and communicate the final decision, planning its implementation to ensure clearly defined roles and responsibilities.Their expertise and facilitation skills are crucial for maintaining a structured, transparent, and effective decision-making process.
Despite their critical role, practitioners face several challenges throughout the MAMCA process.Firstly, setting up the evaluation structure requires careful planning and coordination to ensure all relevant criteria and alternatives are considered.Practitioners must manage diverse stakeholder inputs and preferences, which can be time-consuming and complex.Additionally, facilitating weight elicitation and alternative appraisal involves ensuring stakeholders understand the methods and can effectively contribute, which often necessitates specialized knowledge and tools.Another significant challenge is visualizing aggregated preferences and facilitating consensus-building discussions in a way that is clear and comprehensible to all stakeholders.These challenges underscore the need for effective tools and decision-support systems (DSS) to enhance the decision-making process.
To enhance the efficacy and utility of MAMCA in real-world scenarios, specialised MAMCA software has been developed (Huang et al., 2020).This software serves as an interactive platform, designed to assist practitioners and stakeholders.in navigating complex decision-making landscapes.Figure 1 also depicts the steps that the software can support.The MAMCA software supports the MAMCA process in several ways: 1.After identifying the alternatives, criteria, and stakeholders, the software documents this data in a database, preparing it for further evaluation.
2. Interactive tools within the software facilitate systematic weight elicitation using methods such as the Analytic Hierarchy Process (AHP) (Saaty, 2003) or Direct Rating (Bottomley et al., 2000).This ensures stakeholders can easily understand and participate in the weight elicitation process.3. The software provides modules for evaluating alternatives against the criteria, supporting various MCDM methods, and guiding stakeholders comprehensively through the evaluation process.4. If the process opts for output-level aggregation among stakeholder groups, the software automates the aggregation of preferences and clearly visualizes the results.5.If the process does not opt for output-level aggregation, the software offers a series of visualizations to facilitate consensus reaching.One important feature in MAMCA is the Multi-Actor view (Macharis et al., 2012), which displays the preferences of different stakeholders and highlights potential conflicts.This visualization helps practitioners by clearly presenting stakeholder group preferences, mediating discussions, reconciling conflicts, and guiding stakeholders towards consensus.6.Finally, the software provides sensitivity analysis to validate stakeholder preferences by allowing for the modification of weights.
The interactive nature of the software allows for real-time adjustments and evaluations, thereby fostering a more dynamic and collaborative decisionmaking process.The software makes the workshop setting more manageable, where quick teamwork and getting everyone on board are important and a crucial success factor.It relieves the burden on practitioners during workshops, enabling stakeholders to complete the evaluation process in a semi-autonomous manner.It has supported decision-making processes across different countries and sectors (Brusselaers et al., 2021;Keseru et al., 2021;Lode et al., 2023).Furthermore, the software is undergoing refinements to more effectively align with social decision-making models, thus expanding its versatility across a diverse range of industries and decision-making approaches (Huang et al., 2021).
In the aftermath of global pandemics, the need for hybrid stakeholder involvement in social decision-making has become increasingly evident (Tagliaro & Migliore, 2022;Vyas, 2022).Besides the cause of the COVID-19 pandemic, various factors have complicated stakeholder involvement.For instance, certain stakeholders are geographically dispersed or otherwise difficult to engage, making them ideal candidates for online evaluation methods (Brisendine et al., 2023).For example, in a MAMCA study appraising toll schemes in Australia, the main stakeholders were members of the Transport Association which includes many large and medium-sized transportation and logistics companies (Perera & Thompson, 2021).Over 100 companies were invited to participate in the decisionmaking process.It is difficult for the practitioners to engage this amount stakeholders to the workshop.The authors distributed questionnaires to complete the evaluation, a task that the current MAMCA software could not support at the time.
Another finding was the time-consuming and challenging nature of instructing non-operational research stakeholders to elicit their preferences using MCDM methods (Huang et al., 2021).Due to the complex mathematical background of MCDM methods, such as PROMETHEE (Brans & De Smet, 2016), stakeholders often struggle to quickly grasp how to structure preference parameters properly, even with practitioners' guidance.In some MAMCA cases, preference functions are structured with by experts, as seen in the assessment of energy scenarios in a bioenergy village in Germany (Sch€ ar & Geldermann, 2021).In social decision-making, where most criteria are qualitative, practitioners can use some simple MCDM methods in MAMCA software to provide visual interaction for eliciting preferences.However, stakeholders may still find the methods difficult to understand and time-consuming, as reported in the case of Brusselaers et al. (2021).The software needs to develop more intuitive approaches for preference elicitation, potentially requiring minimal or no guidance from practitioners.
Survey-based decision-making integrated in the software, has emerged as a viable solution to these challenges.This approach allows representatives to participate in the evaluation process either online or offline, thereby increasing the accessibility and inclusiveness of the decision-making process.
Furthermore, survey-based decision-making aligns well with recent advancements in the MAMCA framework.Recent MAMCA cases have shown researchers developing a new evaluation structure within the framework.In studies by Hamadneh et al. (2022); Keseru et al. (2021); Khattak et al. (2022); Perera and Thompson (2021), more than five representatives were invited to represent each stakeholder group.Instead of conducting input-level aggregation within groups and output-level aggregation among groups, each individual expressed their preferences during the evaluation process.Within stakeholder groups, representatives elicited preferences individually, leading to a two-layer output-level aggregation scheme, or output-level aggregation within each group and no aggregation among groups.It actually develops MAMCA into a framework that encompasses multiple layers of group decision-making or groups of group decision-making.This new evaluation structure and aggregation scheme are efforts by researchers to exploit the MAMCA framework's potential to capture more detailed preferences and other characteristics from stakeholders.The survey-based decision-making integration can facilitate this new scheme, making the evaluation structure manageable.
Despite these advancements, achieving a balance between a user-friendly, easily understandable process and attaining robust, merit-based results remains challenging.Therefore, it is essential to first clarify the evaluation structure and then design tailored methods to ensure effectiveness.In the next section, we present the mathematical foundation of the MAMCA evaluation structure with output-level aggregation within stakeholder groups.

Exploring individual preferences within stakeholder groups in the MAMCA framework: Mathematical foundations
The new aggregation scheme we propose requires a detailed investigation of individual preferences within stakeholder groups.Consequently, a welldefined mathematical framework is essential to properly structure these preferences.Such a framework will enhance readers' understanding of the interplay among attributes within MAMCA.
We now focus on the structure of the outputlevel aggregation within each group and no aggregation among groups.After defining the problem and objectives, practitioners are tasked with identifying viable alternatives that could potentially resolve the issue at hand.At the same time, the relevant stakeholder groups that should participate in the MAMCA process are identified, along with the criteria relevant to each group.Let us denote a distinct set of alternatives for evaluation as A ¼ fa 1 , a 2 , :::, a m g: We identify key stakeholder groups that should be involved in the decision-making process, denoted as S ¼ fs 1 , :::, s k g: Each of these groups may comprise various participants (ie, representatives), represented as R i ¼ fr i, 1 , :::, r i, h i g where h i � 1 and i ¼ 1, :::, k: Distinct criteria can be defined for each group, forming the criteria set G i ¼ fg i, 1 , :::, g i, n i g: The criteria can be clustered under identified sub-objectives/dimensions, eg, environmental dimension, economic dimension.Additionally, it is feasible to establish sub-criteria under an individual criterion, expressed as G 0 i, j ¼ fg 0 i, j, 1 , :::, g 0 i, j, n j g for j ¼ 1, :::, n i : This representation underscores that stakeholder groups might differ in the representative numbers, criteria, and subcriteria, with the MAMCA framework showcasing apparent hierarchical tiers, as illustrated in Figure 2. It is important to note that the procedure for the initial phase of the MAMCA process can be adapted to meet specific needs.For example, a comprehensive list of criteria G could be established first.Once the stakeholders have been identified, they would be tasked with selecting the most relevant criteria from this list for their respective analyses (Huang et al., 2023).
The subsequent phase in the MAMCA methodology is recognised as the evaluation preparation phase.At this point, criteria are established that must be quantifiable through one or more indicators, thus allowing the performance of different alternatives to be quantified.This quantification can be based on objective data or subjective judgments.Consequently, the performance of an alternative a i on a given criterion g j can be expressed as g j ða i Þ: It is at this stage that appropriate MCDM method is selected.Various group decision methods that extend traditional MCDM approaches can be used The next stage is evaluation.With the MCDM methods, In the MAMCA approach, representatives are tasked with determining the relative importance of criteria, known as weights.For a given representative r i, j within the stakeholder group s i , these weights can be expressed as W i, j ¼ fw i, j, 1 , :::, w i, j, n i g corresponding to the respective criteria g i, i 0 2 G i : In addition, the representatives are expected to express their preferences for the alternatives with respect to these criteria.The elicitation of these preferences may vary depending on the MCDM method used, which may involve pairwise comparisons, normalisation processes, or other established approaches (Cinelli et al., 2021).We represent the preference score of an alternative with respect to the criterion i 0 as f i 0 ðg i 0 ðaÞÞ, which expresses a quantifiable metric reflecting the representative's preference based on its performance.To derive a representative j's within stakeholder group i overall preference, these individual criterion scores are aggregated, taking into account the associated weights of each criterion: v i, j ðaÞ ¼ f ðff i, j, 1 ðg i, j, 1 ðaÞÞ, :::, f i, j, n i ðg i, j, n i ðaÞÞg, fw i, j, 1 , :::, w i, j, n i gÞ, where the aggregation method f used to combine criteria scores can be different, adapting to the specific context or preference of the decision-making process (Cinelli et al., 2021).Typically, MAMCA uses an additive model for aggregation, which can be represented as follows: The additive model is preferred in the MAMCA framework because it promotes ease of understanding among stakeholders and has broader applicability compared to alternative approaches such as geometric or harmonic means, which can be only applied in strictly positive normalized scores (Gasser et al., 2019).Moreover, the linear expression of the additive model lays the groundwork for potential optimization procedures aimed at identifying solutions that achieve consensus among stakeholders (Huang et al., 2021).The individual preferences expressed by the representatives can then be consolidated into a collective preference for the stakeholder group.In situations where the goals of the representatives are aligned, the group preference can be computed simply by taking the arithmetic mean of the individual evaluations, denoted as V i ðaÞ ¼ 1 h i P h i j¼1 v i, j ðaÞ: Then, stakeholder group preferences for the various alternatives are collected, resulting in the construction of a preference matrix: The preference matrix (3) is not intended to be further aggregated to derive a single global preference for alternatives in the scheme we discussed.Due to the potentially divergent objectives of different stakeholder groups, it is more important to investigate the preferences of the stakeholder groups without aggregating.In practice, MAMCA advocates discussion among stakeholders to reach a consensus, which marks the transition to the next phase of MAMCA, which involves discussion of results and consensus building (Huang et al., 2021).The preference matrix (3) can be visualized using a multi-line graph -referred to as the multi-actor view -to simultaneously display the alternative preferences of all stakeholder groups (Huang et al., 2020;Macharis et al., 2009).This graphical representation facilitates mutual understanding among stakeholder groups by making clear each group's preferences and positions.It serves for representatives from different groups to engage in discussion, express their perspectives, debate the pros and cons of the alternatives, and seek a consensus solution.The discussion can be facilitated by a consensus model (Huang et al., 2021).
Finally, stakeholder groups may select a final alternative for implementation, or they may initiate a subsequent round of evaluation if changes to the current alternatives are needed or new alternatives emerge.This iterative process ensures that decisionmaking remains dynamic and adaptive to changing conditions or insights, allowing for continuous refinement of alternatives until the most favourable outcome is selected for action.
A pseudocode algorithm can help readers to better understand the MAMCA process, as shown in follows Algorithm 1.
Algorithm 1. MAMCA decision-making process with the structure of the output-level aggregation within stakeholder groups Input: A decision-making problem to address.

end for
Aggregate representatives preferences Collect and illustrate stakeholder groups' preferences Discuss and negotiate A solution to the decision-making problem, ie, a selected :

The survey-based MAMCA DSS
Based on the structure presented, it is evident that a thorough understanding of stakeholders' preferences is essential.As mentioned in Section 3, integrating a survey-based approach into the MAMCA software can significantly facilitate this process.This integration simplifies stakeholder engagement, allowing for the inclusion of larger groups of participants and enabling them to provide detailed input.The survey-based approach also enables effective output-level aggregation within stakeholder groups.By collecting individual preferences through structured surveys, the software can systematically aggregate these preferences to reflect the collective opinion of each group.This method provides a clear and transparent mechanism for capturing and combining diverse viewpoints, ensuring that the final decisions are representative of the entire group.
The survey-based approach ensures that stakeholders can conduct effective evaluations even in the absence of direct, face-to-face guidance from practitioners, provided they receive proper instructions.It enhances the flexibility and scalability of the MAMCA framework.In a workshop setting, these approaches allow stakeholders to participate both in-person and remotely, accommodating diverse schedules and geographic locations, which broadens the scope of participation.This method also reduces the logistical complexities associated with organizing large-scale, in-person workshops.By leveraging digital tools and platforms, the surveybased approach ensures that data collection is systematic and comprehensive, capturing a wide range of stakeholder preferences.The use of clear, userfriendly interfaces in the MAMCA software further aids stakeholders in understanding and completing the evaluation tasks, thereby improving the quality and reliability of the input received.This digital facilitation is particularly advantageous in contexts requiring quick and adaptive decision-making processes.
However, designing intuitive and comprehensive approaches remains crucial to maximizing the effectiveness of this method.To address this need, we have proposed two adapted survey-based approaches: a hybrid weight elicitation method utilizing the surveybased Simos method and the survey-based Simple Multi Attribute Rating Technique (SMART) for alternative preference elicitation.These adaptations are specifically designed to streamline the evaluation process, ensuring that stakeholders can participate in the decision-making process with greater ease and clarity.By simplifying the weight elicitation and preference evaluation processes, these methods enhance the overall efficiency and inclusivity of the decision-making framework.

Hybrid weight elicitation method
We propose to use a hybrid method to elicit criteria weights.In principle, the revised Simos method will be applied for single-level criteria.However, if there are multiple levels of criteria, ie, hierarchical structure, we will apply direct rating (DR) as a complementary method (Bottomley et al., 2000).The original Simos method, rooted in a card-play game, facilitates weight elicitation for decision criteria (Simos, 1990).Decision-makers are given a set of cards, each representing a criterion, along with a set of white cards.The cards are arranged in an order that reflects the relative importance of the criteria.White cards may be interspersed to indicate a larger perceived gap between adjacent criteria.Equally important criteria can be grouped together as ex aequo sets, either by clipping them together or arranging them adjacently on a table or board.
In an extension to Simos' original methodology, Figueira and Roy (2002) introduced the ability for decision-makers to specify a ratio factor z, which characterizes the importance of the most important criterion relative to the least important one.Let u ¼ z−1 e−1 , where e denotes the total ranks, including white cards (each white card represents one rank).Assume that a criterion g k 2 G is ranked at the j th position, making it the e − 1 − j least important criterion.The non-normalized weight v j for this criterion is then calculated as follows: (4) To normalize the weights to the interval [0,1], each criterion's weight w j is obtained by dividing its non-normalized weight by the sum of all non-normalized weights, which ensures that the sum of all criteria weights equals one, ie, P n i¼1 w j ¼ 1: While the revised Simos method offers an intuitive approach for weight elicitation, implementing an interactive card-play game in a survey setting remains challenging.To this end, our study introduces a point-scale variant of the revised Simos method in the purpose of a survey design.
We define a 5-point scale, where participants assign scores ranging from 1 to 5 to different criteria.We instruct participants that 5 represents the highest importance for the criteria, while 1 signifies the least importance.At least one criterion must receive a maximum score 5.The purpose of this decision is to maintain consistent options for participants to select a value for the variable, z.Although the scale used is ordinal, participants are instructed that the lowest score of l indicates their assumption that the most important criteria are 5=l times important as the least important criteria.Unassigned scores between the least and most important criteria are treated analogously to white cards in the original revised Simos method.Accordingly, the non-normalised weight x 0 i for a criterion scored at i points is calculated as: where u ¼ 5=l 5−l : After normalization, this allows us to obtain the criteria weights for one participant.Suppose we have a set of participants P ¼ fp 1 , p 2 , :::, p n g: For each participant p i , we derive a vector of criteria weights W i ¼ fw i, 1 , w i, 2 , :::, w i, m g using the point-scale revised Simos method as described earlier.Combining the weight vectors from all participants, we can construct a matrix of participant-specific criteria weights, W, defined as follows: In scenarios with a single level criteria structure, our methodology is straightforward.However, when dealing with hierarchical structures involving multiple levels of criteria, we shift to applying Direct Rating (DR) to the lower levels.This shift is particularly relevant when the number of subcriteria under a main criterion is relatively small, usually in the range of 3-5.In these cases, the Simos method becomes less effective, while direct rating becomes a more appropriate method for determining the importance of each criterion.
Participants are asked to rate the criteria on a 5-point scale.For example, consider a criterion g i with n i subcriteria.A participant i 0 assigns a rating to the importance of a subcriterion g 0 i, j , denoted as l 0 i, j : The weights of the subcriteria are then calculated as follows: where, w 0 i 0 , i, j represents the calculated weight of the subcriterion g 0 i, j for the participant i 0 : The local weight, derived by DR, is multiplied by the main criterion weight w i 0 , i : This formulation ensures a comprehensive weighting system that accurately reflects both the hierarchical structure of the criteria and the relative importance of each subcriterion.

Survey-based SMART method for alternative preference elicitation
The Simple Multi-Attribute Rating Technique (SMART), a compensatory approach in multiple criteria decision making, was developed by Edwards (1977), and originally designed as a simplified implementation of the Multi-Attribute Utility Theory (MAUT) (Figueira et al., 2005).Edwards and Barron identified and addressed certain limitations in the original SMART method, leading to the development of two improved variants: SMARTS and SMARTER (Edwards & Barron, 1994).These three methods use different weight methods, but share the same way to calculate the alternative scores based on their performances on criteria.Thus, we can apply the same logic to our appraisal approach.One of the most notable advantages of the SMART method lies in the simplicity of its questioning and response mechanism.This simplicity significantly aids in enhancing the decisionmaker's understanding of the process, thereby facilitating a more intuitive approach to solving the problem.In our research, we applied the SMART method in a survey format, aiming to provide an accessible and user-friendly method for eliciting stakeholder preferences, making it more approachable for stakeholders who may not be familiar with complex MCDM techniques.
With the weights already elicited using the proposed method, our focus shifts to employing the SMART method specifically for eliciting preferences for alternatives.This is achieved by asking participants to assign scores to each alternative based on different criteria.In the original SMART methodology, the elicitation of preference or performance of an alternative a i , in relation to a specific criterion g i 0 , is determined based on the direction of the indicator: where g i 0 ð�Þ denotes the scores of criterion g i 0 measured by corresponding indicator.q i 0 , max and q i 0 , min denote the maximum and minimum plausible values for criterion j, respectively.Importantly, these values are not strictly the boundaries of the alternatives' performances but rather the feasible range within which the alternatives may vary.In our adapted survey setting, each indicator is transformed into a 10point scale, to simplify the evaluation process for participants.Within this scale, a rating of 0 represents the lowest possible preference for an alternative with respect to a specific criterion, while a rating of 10 signifies the highest possible preference.
Participants are requested to assign a score within this range for each alternative g i 0 ða i Þ, based on their assessment of its performance against each criterion.Again the evaluation of a i on criterion g i 0 will be: In the proposed approach, we essentially normalize the scores to a base value of 1, making the conversion to a 100-point scale redundant.This normalization simplifies the scoring process and maintains the clarity of the evaluation.With these methods, participants can effectively complete their evaluations online, independently, and without the need for direct guidance from practitioners.
At the end, the final preference of the alternatives can be aggregated by the additive model ( 2), where the weights are obtained based on the proposed hybrid weight elicitation method, the preferences are elicited by SMART.

Extend MAMCA DSS with the survey-based approaches
To streamline the workflow of the survey-based MAMCA for practical applications, we have integrated the aforementioned survey functionalities within the MAMCA software 1 .This integration aims to enhance the flexibility, scalability, and effectiveness of the decision-making process by leveraging digital tools and platforms, making the new aggregation scheme feasible within the software.The new extension has the following features: � User-friendly interface and accessibility.
Stakeholders can complete evaluations through both the weight elicitation survey and the alternative appraisal survey directly within the software without the need to log in; scanning a QR code is sufficient for access.This accessibility feature reduces barriers to participation and ensures that stakeholders can easily engage with the decision-making process.The software's interface is designed to be intuitive and user-friendly, facilitating a smooth and efficient evaluation process.� Survey introduction and instruction.Figure 3   illustration of the alternatives and criteria to be evaluated.It ensures that participants have a clear understanding of the options and the metrics by which they will be assessed.� Survey-based preference elicitation process.For the weight elicitation process, participants complete the questionnaire depicted in Figure 3(c).
Using the data gathered from these surveys, weights are elicited by employing the hybrid weight elicitation methodology detailed in Section 4.1.For the alternative appraisal, participants record their responses in the questionnaires shown in Figure 3(d).This process uses the survey-based SMART method described in Section 4.2.� Systematic data collection and analysis.The survey-based approach ensures that data collection is systematic and comprehensive, capturing a wide range of stakeholder preferences.Practitioners can easily access individual representatives' preference results within the software.This capability enhances the overall transparency and understanding of each component of the evaluation as illustrated in Figure 2. � Real-time feedback and dynamic interaction.
During workshops, stakeholders can use the digital tools provided by the MAMCA software to complete surveys and evaluations in real-time.This feature promotes immediate feedback and dynamic interaction, allowing for quick adjustments and continuous engagement with a larger and more diverse group of participants.This digital facilitation is particularly advantageous in contexts requiring quick and adaptive decisionmaking processes.
In summary, the integration of survey-based approaches within the MAMCA DSS significantly enhances the overall efficiency and inclusivity of the decision-making framework.By simplifying the weight elicitation and preference evaluation processes, these methods ensure that stakeholders can participate in the decision-making process with greater ease and clarity.This development is particularly crucial for implementing the new aggregation scheme proposed in the previous section, which involves output-level aggregation within stakeholder groups.The integration not only maintains the interactive and collaborative nature of traditional workshops but also leverages the advantages of digital facilitation, thereby improving the transparency and effectiveness of the decision-making process.

Real-life case demonstration: Swiss energy policy evaluation
To showcase its capabilities, we employ the surveybased DSS in a real-life case study to evaluate Swiss energy policy in the socio-political dimensions.Switzerland is a country with a strong commitment to climate neutrality and energy security.However, the country is not on track to meet its goals and is facing a number of challenges in achieving these goals.In order to overcome future challenges, an energy policy framework is needed that enables domestic electricity production to be expanded as quickly as possible, extensive electrification/decarbonisation of the energy supply, and increasingly efficient use of energy.Progress in Switzerland is also influenced by fundamental changes in the energy markets.These are caused by economic, political, and technological developments at home and abroad.Among these, the socio-political aspects represent a critical aspect that requires in-depth analysis.The socio-political dimension encompasses the variety of interests, preferences, and values held by different stakeholders.Understanding how sociopolitical criteria interact with policy decisions is essential for developing effective and socially acceptable strategies for Switzerland's energy transition.The diverse involvement of stakeholders, coupled with the qualitative nature of the assessment of socio-political criteria, makes this a suitable case to exemplify the application of survey-based MAMCA.
In this study, we adopted a role-play approach, engaging scientists, PhD candidates, and Master's students specialising in energy systems based in Switzerland to undertake a comprehensive evaluation of the policies.These participants were assigned to represent four distinct stakeholder groups: s 1 Government and Regulatory Authorities, s 2 Energy Providers and Utility Companies, s 3 Financial Institutions and Investors, and s 4 Consumers and Civil Society.For illustrative purposes, our analysis concentrates exclusively on subsidy-type policies.As detailed in Table 1, we have identified five such policies currently enacted at either the national or cantonal level in Switzerland.Additionally, we examine policies known as "Mantelerlass" (overarching decree), designed to steer Switzerland's energy supply system towards the net-zero emissions target by 2050 while maintaining a high level of supply security.Furthermore, we consider the establishment of a State Investment Bank (SIB)-A successful model in Germany-recommended by researchers for adoption in Switzerland (Egli et al., 2022).The development of our seven criteria, as outlined in Table 2, is based on an extensive literature review.
The aforementioned structure of MAMCA is implemented within the MAMCA software by following established guidelines (Huang et al., 2020).To validate the effectiveness of the survey-based decision-making process, the role-play subjects are as follows: � Remote decision-making validation: To test the new DSS's effectiveness in a remote decision-making context, the evaluation is not conducted in a workshop setting but rather in an individual setting.Participants had bilateral meetings with practitioners before the evaluation.As practitioners and facilitators, we introduced the background and objective of the decision-making problem.We provided basic knowledge of MCDM to ensure participants understood the MCDM evaluation process.However, to test the clarity of the survey introduction and instruction in the DSS, we did not disclose the exact alternatives and criteria, and the survey-based method details. � Role-play notification and selection: Participants were informed that they would evaluate alternatives as stakeholders in a role-play scenario.The first several interviewed participants were allowed to choose the stakeholder roles they felt confident in.Interviews were conducted in order of experience: master's students first, followed by PhD students, and then scientists.This approach allowed less experienced participants to select roles they were more comfortable with.For example, a master's student who felt less confident in the roles of s 2 Energy Providers and Utility Companies, and s 3 Financial Institutions and Investors, chose to play the role of s 4 Consumers and Civil Society.Conversely, a PhD student specializing in energy economics chose the role of s 2 Energy Providers and Utility Companies.� Survey link distribution: After selecting their stakeholder roles, participants received a dedicated survey link, which they could follow to undertake the survey-based decision-making process, as illustrated in Figure 3.The meeting concluded, and participants were able to complete the evaluation individually.� Post-evaluation interview: Upon completing the evaluation, participants had individual interviews where they could review the results.They were asked if the results corresponded to their perceptions and to provide additional feedback, such as the ease of use of the new survey functions and any potential issues encountered during the evaluation.
By following this procedure, we aim to ensure that the survey-based decision-making process is effective, clear, and user-friendly for all participants.The participants follow the online pages to do the evaluation in the MAMCA software.For each stakeholder group, 2 participants are involved.After the evaluation, practitioners are also able to review the overall results at a high level.The group result is illustrated in Table 3.It is also possible to check the results of stakeholder groups and at the individual level.
Given the focus of this study on the development of the DSS and the case serves as a demonstration, we demonstrate the functionality of the system rather than delving into an extensive analysis of casespecific results.The DSS offers a range of data visualisations for comprehensive analysis (see Figure 4).To understand the overarching preferences of different stakeholder groups, one can refer to the multi-actor  For a more detailed view of alternative performance, the group bird's-eye view can be checked; for example, to examine the scores of s 1 Government and Regulatory Authorities, one can refer to visualizations such as those shown in Figure 4(b).These visualizations show the scored preferences for each criterion alongside the final scores for the alternatives, with the weights displayed as a bar chart and the criteria ranked by importance -providing clarity on stakeholder priorities.Our analysis reveals that the criteria of" Public Support" and" Policy Effectiveness" are deemed highly important.Notably, alternative a 3 scores highly across the majority of criteria, establishing it as the most favored option for this stakeholder group.
Should stakeholders wish to examine the priority homogeneity within their group's responses, the average result view is available.It uses box plots to convey the homogeneity or heterogeneity in the importance that participants within a stakeholder group place on various criteria.For instance, examining the average results for s 3 Financial Institutions and Investors, Figure 4(c) reveals that the weights assigned to various criteria exhibit a high degree of homogeneity.Only minor variations in the weights attributed to different criteria are observed.Also, practitioners have the ability to delve into individual participants' scores with charts similar to the bird's eye view shown.They can further interact with the data by adjusting criterion weights directly through the bar graph.A simple sensitivity analysis, Weight Stability Intervals (WSI), is applied to assess the robustness of individual results (Mareschal, 1988).

Discussion
After completing the evaluation and obtaining preferences from all participants, we conducted follow-up interviews to gather feedback on the surveybased MAMCA DSS.
We identified three different types of participants to collect varied feedback: 1. Participants who have never been involved in the MCDM evaluation process and have no knowledge of MCDM. 2. Participants who have previously used MCDM or have been involved in MCDA evaluation processes.3. Participants who have used previous versions of the MAMCA software.
For participants with no prior MCDM experience, we asked if the designed survey-based MCDM methods accurately reflected their preferences and if the methods were easy to understand.Overall, these participants found the methods intuitive and the instructions clear.One participant highlighted the option to mark criteria as irrelevant.Generally, they agreed that the elicited weights represented their preferences.Regarding alternative preference elicitation, they found the survey-based SMART method straightforward and logical.Overall, they did not encounter difficulties while completing the survey.
For participants with previous MCDM experience, we asked additional questions to compare the survey-based methods with other MCDM methods they had used before.This comparison helped us understand the relative effectiveness and userfriendliness of our approach.Participants found the survey-based methods easier to navigate compared to traditional MCDM tools.They reported that the survey-based approach streamlined the evaluation process, allowing them to complete assessments more quickly without sacrificing accuracy.One participant, who had experience applying AHP to elicit preferences, noted that the point-scale evaluation was more straightforward and efficient.Given the number of criteria and alternatives, he expressed confidence that the survey-based approach could generate robust results that accurately reflected his preferences.Overall, participants felt that the survey-based methods adequately captured their preferences, comparable to or better than other MCDM methods they had used.The digital format and interactive nature of the surveys kept participants more engaged throughout the evaluation process.One participant mentioned that she could easily switch between pages, which helped her refresh the provided information and review her inputs more effectively.This feature contributed to a more seamless and interactive evaluation experience.
For participants who had used previous versions of the MAMCA software, we inquired about their perceptions of the differences between the survey-based MAMCA and the conventional MAMCA process in the software.In our case, only one participant had prior experience with MAMCA.He found the new survey-based functions to be more user-friendly, with a more intuitive interface that made it easier to navigate and understand the process.The integration of survey-based approaches streamlined the overall workflow, reducing the time and effort required to complete evaluations while maintaining or improving the quality of the outputs.He highlighted that in his previous MAMCA software experience, he needed to follow instructions from practitioners due to the lack of built-in guidelines.However, with the survey-based MAMCA, he was able to complete the elicitation process independently by following the guidance provided within the software.Additionally, the ability to access and complete evaluations remotely via a simple QR code scan was seen as a major improvement, particularly in facilitating participation from a broader range of stakeholders.The interactive elements of the surveys kept participants more engaged compared to the traditional MAMCA process, resulting in a more active and involved decision-making experience.When informed that the objective of developing this survey-based approach was to facilitate the outputlevel aggregation evaluation structure within stakeholder groups, he agreed that this approach indeed facilitated the aggregation scheme.It became easier to engage more participants within stakeholder groups, and the software made it possible to check the results from different participants easily.Regarding data visualization, the participant appreciated the new integrated group bird's-eye view, which made it easy to check the preferences of alternatives on criteria with priority order.He believed the intra-group view could facilitate result discussions within stakeholder groups by illustrating the priority differences among participants within the same group.This feature is important for an output-level aggregation scheme, as it enhances the transparency and clarity of the decision-making process.
In conclusion, by incorporating survey-based weight elicitation and survey-based alternative appraisal functions, along with multiple data visualizations, the DSS provides an innovative approach to decision-making from multiple stakeholder perspectives.This approach is particularly well-suited for both online and offline workshops, ensuring extensive participant engagement and flexibility.The ability for stakeholders to complete questionnaires without logging in simplifies the evaluation process, making it more accessible and userfriendly.The MAMCA software's capability to facilitate output-level aggregation within stakeholder groups is a significant advancement.This feature enables a comprehensive group-of-group decision-making process by allowing for the detailed collection and aggregation of individual preferences within each stakeholder group.Practitioners can easily present results at various levels, providing a clear and transparent overview of the aggregated preferences.This helps in identifying consensus and differences within and across groups, thus supporting more informed and balanced decision-making.Moreover, the software's interactive elements and integrated data visualizations, such as the group bird'seye view and intra-group priority comparisons, enhance the clarity and comprehensibility of the decision-making process.These visual tools allow stakeholders to see the impact of their inputs and the collective preferences of their groups, fostering a deeper understanding and more meaningful engagement.Overall, this DSS significantly improves stakeholder engagement and promotes transparency throughout the decision-making process.It ensures that the diverse preferences of stakeholders are systematically captured, aggregated, and presented.The system's flexibility and user-friendly design make it an invaluable tool for complex decision-making scenarios involving multiple stakeholders.

Limitations and future work
In this study, we have developed survey-based MCDM methods within the MAMCA software, extending it into a survey-based DSS designed to facilitate output-level aggregation within stakeholder groups.By demonstrating the survey-based DSS in a case study, we identified several advantages of this new approach compared to the previous version of the MAMCA software.However, we also noted several limitations during the implementation of the proposed approach.
First, while the survey-based MCDM methods are intuitive and easy to understand, they are currently more suited for qualitative indicator evaluation, as demonstrated in the case study evaluating Swiss energy policies with socio-political criteria.It is necessary to validate these methods in cases involving quantitative indicators.In such instances, integrating other preference elicitation methods, such as PROMETHEE, could be considered to capture detailed preferences from quantitative data.
Given that the survey-based DSS is designed to facilitate output-level aggregation, it is crucial to engage more participants within groups to validate its effectiveness in reflecting individual preferences in larger participation.Currently, we assume that representatives from the same groups hold similar preferences, allowing us to aggregate their preferences by averaging their results.However, this assumption may not hold in cases of mass-participation, where hundreds of participants are involved (Huang, 2023b).A systematic method for assessing homogeneity or heterogeneity within these groups is currently absent.This gap becomes particularly significant as survey-based evaluation techniques allow for the participation of an increasing number of stakeholders.Developing robust methodologies to accommodate and analyze the complexities introduced by larger, more diverse participant pools is essential.
The current survey-based approach focuses on facilitating the evaluation process.Although the tool supports both workshop and non-workshop settings, there is a need to enhance stakeholder interaction in non-workshop settings.The decision-making process encompasses more than just evaluation.As the MAMCA framework aims to engage stakeholders at a higher level, rather than merely involving them in the evaluation as a form of consultation (Davidson, 1998), further development is needed to involve stakeholders in later stages, particularly in the consensus-reaching process (CRP) (Zhang et al., 2019).
Therefore, it is also crucial to enhance the DSS's capacity for nuanced post-hoc analysis, such as CRP, which is a critical part of decision-making within the MAMCA framework.CRP should identify possible compromise solutions for stakeholders.Previously, a consensus-reaching model based on weight modification showed considerable promise in addressing the challenges posed by divergent preferences and objectives among stakeholder groups (Huang et al., 2021).However, full integration of this model into the MAMCA DSS has yet to be achieved.Incorporating the consensus-building function into the DSS will greatly enhance its utility in facilitating more effective and harmonious decision-making processes.
Addressing these limitations and areas of future work will not only enhance the depth and breadth of analysis possible with the MAMCA DSS but will also ensure that the system evolves to meet the increasingly complex and multifaceted nature of stakeholder-based decision making.By refining consensus-building capabilities and deepening the understanding of intra-group dynamics, the MAMCA framework will be better positioned to support comprehensive, inclusive, and effective decision-making processes across various domains.

Conclusion
In this study, we have extended the capabilities of the MAMCA software by integrating survey-based MCDM methods, transforming it into a more accessible and inclusive DSS.This extension is specifically designed to facilitate the output-level aggregation scheme within stakeholder groups, allowing for detailed collection and aggregation of individual preferences.The incorporation of survey-based techniques such as the survey-based Simos method and the survey-based SMART method streamlines the weight elicitation and alternative appraisal processes, making them intuitive and easy to understand for a wide range of stakeholders.
The effectiveness of this survey-based apporach was demonstrated through a case study evaluating Swiss energy policies, showcasing its ability to support comprehensive stakeholder assessments and facilitate understanding across diverse groups.The new survey-based DSS offers significant advantages over the previous version, including increased scalability, flexibility, and the ability to engage more participants, both online and offline.It simplifies the evaluation process by allowing stakeholders to complete questionnaires without logging in, thereby reducing barriers to participation and ensuring that diverse preferences are systematically captured and aggregated.
Despite these advancements, several limitations and future work directions were identified during the implementation of the proposed approach.In conclusion, the survey-based MAMCA DSS represents an advancement in facilitating comprehensive and inclusive stakeholder-based decision-making processes and deepening the understanding of intragroup dynamics.

Figure 1 .
Figure 1.MAMCA workflow aligned with the generic MCGDM framework process.Icons represent the involvement of practitioners, stakeholders, experts, and software at various stages, where the red dashes on the icons indicate their potential involvement but not mandatory participation.
, including the potential to combine different methods.Examples of such combinations include the Preference Ranking Organisation Method for Enrichment Evaluation (PROMETHEE) and the Analytic Hierarchy Process (AHP) (Macharis et al., 2012), the Simple Multi-Attribute Rating Technique (SMART) with AHP (Huang et al., 2020; Lode et al., 2022), Simos method and PROMETHEE (Huang et al., 2023) and the integration of the Best Worst Method (BWM) with PROMETHEE (Huang, 2023a; Sivakumar et al., 2020).
j ðaÞ, 8a 2 A end for Stage 4: Result discussion and consensus reaching illustrates the overview pages of the survey.In both the weight elicitation and alternative appraisal surveys, participants are first presented with two informational pages before beginning the evaluation: � Decision problem introduction (Figure 3(a)): This page provides an introduction to the decision problem at hand, along with an indication of the role of the stakeholder group.It ensures that participants understand the context of their input.� Alternatives and criteria overview (Figure 3(b)): This page offers a comprehensive

Figure 3 .
Figure 3. Screenshots of the questionnaires.
Identify stakeholder groups S ¼ fs 1 , :::, s k g for s i in S do .1 st round of stakeholder workshop (Potential) Define criteria for group i: G i ¼ fg i, 1 , :::, g i, n i g end for for g in G i , i ¼ 1, 2, :::, k do Stage 1: Problem structuring Identify a set of alternative A ¼ fa 1 , a 2 , :::, a m g

Table 1 .
Selected subsidy type Swiss energy policies.

Table 3 .
Alternative preferences from different stakeholder groups.