Participatory Design Study

• Participatory Design Study is defined as a structured empirical investigation where stakeholders collaboratively co-create and refine technological solutions.
• The study employs diverse techniques such as workshops, interviews, prototyping, and live enactments to capture latent needs and design requirements.
• Key insights highlight democratized decision-making, effective power-sharing, and context-specific frameworks that drive robust technology design.

A Participatory Design Study is a structured empirical investigation in which diverse stakeholders co-create, critique, and iteratively develop technological artifacts, systems, or processes. These studies, grounded in the principles of Participatory Design (PD), integrate end-user, expert, and facilitator perspectives to surface latent needs, elicit design requirements, generate artifacts, and drive strategic design decisions through direct involvement and shared agency. Methodologies, participant recruitment, data collection, analysis techniques, and outcomes are tailored to the domain-specific context and research questions, ranging from the design of socially assistive robots to frameworks for AI, public participation in civic spending, or interface specification in high-stakes legal applications.

1. Definitional Foundations and Scope

Participatory Design studies are characterized by close, structured stakeholder involvement throughout the design process, centering on joint exploration, idea generation, prototyping, and iterative feedback. PD is not merely end-user testing but rather a fundamentally democratic process in which all relevant stakeholders—end users, experts, facilitators, and sometimes indirect stakeholders—act as co-designers or co-inquirers. Key traditions framing PD studies include Scandinavian participatory design, co-design, value-sensitive design, participatory action research, and service design, each emphasizing variable degrees of agency, scope of participation, and power-sharing (Wacnik et al., 2024, Delgado et al., 2023, Delgado et al., 2021).

Dimensions analytically used to structure participatory studies include: the motivation for PD (instrumental vs. normative), the scope and timing of agency (what design decisions are open to influence), participant selection (who participates and how), the forms of engagement (workshops, surveys, prototyping, deliberation), and explicit mapping of power relations and decision rights (Delgado et al., 2023, Delgado et al., 2021).

2. Methodologies and Participatory Structures

PD studies utilize a diverse toolkit of participatory techniques—often in combination—to achieve both breadth and depth of stakeholder engagement. A systematic review found that 90% of PD articles employed two or more participatory techniques, most commonly participatory workshops (73%), stakeholder interviews (69%), and prototyping activities (59%) (Wacnik et al., 2024).

Illustrative methodological components:

• Multi-stage workshop series: e.g., Atom–Synthesis–Fusion cycles, progressing from idea generation to integrative concept prototyping and iterative refinement (Ahmed et al., 2023).
• Structured interviews and focus groups: Used to elicit tacit routines, latent needs, unmet desires, and attitudes toward technology, often serving as the foundation for subsequent artifact-driven activities (Bulgaro et al., 2022, Lalwani et al., 30 Nov 2025).
• Live or Wizard-of-Oz in-situ enactment: Real-time, situated prototyping in the user's context to surface obstacles and allow rapid iteration (e.g., Situated Participatory Design with older adults and robots) (Stegner et al., 2023).
• Scenario-based and artifact-driven externalization: Use of scenario cards, 2D/3D toolkits, role-play, storyboarding, and digital platforms to anchor brainstorming and clarify decision points (Ahmed et al., 2023, Bulgaro et al., 2022, Kopeć et al., 2019).
• Crowdsourcing and online platforms: For large-scale, distributed idea generation and preference elicitation in settings such as eHMI design for autonomous vehicles (Cumbal et al., 23 Jun 2025).
• Participatory simulation: Immersive, role-based interactive simulations (e.g., TREC Legal Track) to embed domain experts and system designers as reciprocal agents in high-stakes task design and evaluation (Delgado et al., 2022).
• Decentralized, community-led PD: Distributed, grassroots co-design as in Queer in AI or abolitionist AI studies, leveraging open communication channels and intersectional governance (QueerInAI et al., 2023, Wang et al., 8 Oct 2025).
• PD in asynchronous and open-source contexts: Integration of participatory tools into continuous, distributed development workflows, as with open source software projects (Hellman et al., 2021).

Recruitment strategies span targeted demographic sampling, open calls, organizational volunteers, expert selection, and hybrid models, adjusting to project scale, context, and equity considerations (Wacnik et al., 2024).

3. Data Collection, Analysis, and Artifact Synthesis

Data sources in PD studies include physical and digital artifacts (e.g., sketches, storyboards, mock-ups, wireframes, Post-it clusters), observational and ethnographic notes, audio/video recordings, survey instruments, and system logs. Analysis is typically performed through open and axial coding (thematic analysis), affinity diagramming, and, where appropriate, content analysis and quantitative methods (e.g., Likert-scale satisfaction, task timings).

Formal measures such as inter-rater agreement (Cohen’s κ) are occasionally applied, though many studies remain qualitative and artifact-centric. In some domains, explicit mathematical or computational frameworks capture design requirements or evaluation metrics; for instance, the TREC Legal Track used F-measure and negative-pool sampling protocols for AI system evaluation (Delgado et al., 2022), while participatory budgeting studies deployed combinatorial optimization and welfare metrics to assess aggregation rules (Fairstein et al., 2023).

The outputs of PD studies include design space models (e.g., mapping robot capabilities to elicited emotions via set-theoretic relations), prototype artifacts, evaluation frameworks (e.g., abolitionist LLM constitutions), guidelines, and documented process intermediaries for replication and benchmarking (Bulgaro et al., 2022, Wang et al., 8 Oct 2025).

4. Key Findings and Case Study Examples

PD studies consistently surface both domain-specific and generalizable insights:

Context/Domain	Key Outcomes/Frameworks
SARs in the Wild (Ahmed et al., 2023)	5-part decision-making framework: social adaptation, mishap management, safety, unpredictability, trust calibration
IoT for Digitally-Hesitant Users (Zallio et al., 2020)	Identification of digital literacy barriers; PD-generated modular, open-source learning materials
eHMI Design (Cumbal et al., 23 Jun 2025)	Strong user preference for visual, multi-modal, directional communication; crowdsourced participatory artefact pool
Productivity SAR for Students (Lalwani et al., 30 Nov 2025)	Co-designed ethical and hardware/software guidelines, stakeholder-vetted through interview+workshop hybrid PD
EHR Components in Healthcare (Robert et al., 26 Mar 2025)	Novel “relational spaces” framework, iterative Telling–Making–Enacting cycle, power rebalancing via artifact & role
Multi-Agent Simulation for Transport (Chen et al., 2009)	Embedded domain-expert and practitioner interaction via agent-user interfaces; log-driven online learning
Legal AI (TREC Legal Track) (Delgado et al., 2022)	Bidirectional, simulation-driven PD with legal experts as topic authorities, precision/recall trade-offs aligned to domain risks

These studies highlight that explicit, multi-stage participatory methodologies yield richer, more robust, and contextually adaptive designs, driving not only artifact creation but also process and power structure innovation.

5. Challenges, Limitations, and Power Negotiation

PD studies face persistent challenges:

• Scope and depth of agency: Many PD efforts are consultative, rarely granting stakeholders control over scope-defining or system-level design decisions. Empirical reviews show the vast majority limit participant influence to UI or feature tuning; only a minority involve stakeholders throughout the full AI lifecycle or grant adjudicative authority (Delgado et al., 2023, Delgado et al., 2021).
• Recruitment and representation: Ensuring diversity, depth, and continuity of stakeholder involvement is demanding. Structural inequities and practical barriers (e.g., language, platform accessibility, geographic/time constraints) are recurrent obstacles (QueerInAI et al., 2023, Wacnik et al., 2024).
• Process opacity and power asymmetry: Project leads, coordinators, or core organizers often retain process control, diluting claimed decentralization or collaboration (Delgado et al., 2023, QueerInAI et al., 2023).
• Scalability and sustainability: High-touch, in-person, or in-situ PD cannot scale effortlessly; hybrid (online/offline), crowd-sourced, and infrastructured approaches offer partial solutions but introduce new challenges (Cumbal et al., 23 Jun 2025, Wacnik et al., 2024).
• Tension between tokenism and genuine agency: Both the “participation washing” critique and empirical findings indicate that procedural participation may conceal underlying power imbalances or limit real stakeholder impact (QueerInAI et al., 2023, Wang et al., 8 Oct 2025).

6. Best Practices, Guidelines, and Emerging Paradigms

Analysis of recent literature and empirical PD studies distills several actionable best practices:

• Engage stakeholders throughout all design phases: The majority of rigorous studies recruit stakeholders iteratively from problem framing through to prototyping, evaluation, and even post-deployment adjustment (Wacnik et al., 2024).
• Employ mixed techniques: Combining participatory workshops, prototyping, context-specific activities, and both direct and indirect engagement emerges as the dominant praxis (Wacnik et al., 2024).
• Balance predetermined and emergent process: Responsive adaptation to stakeholder feedback correlates with increased design equity and acceptance (Wacnik et al., 2024).
• Explicitly structure power-sharing: Deploy mechanisms such as rotating facilitation, transparent charters, shared artifact creation, and decision logs to “flatten” expertise and redistribute design authority (Robert et al., 26 Mar 2025, Delgado et al., 2023).
• Integrate context and lived experience: In-situ enactment and scenario design, as in Situated PD (Stegner et al., 2023), reveal latent needs and deployment barriers unapproachable by abstract ideation alone.
• Support infrastructuring: Sustaining long-term participation beyond project timelines requires training, hand-offs, and community management capabilities (Wacnik et al., 2024).
• Document and report process decisions: Systematic mapping of “Why, What, Who, How, Power” (as in (Delgado et al., 2021, Delgado et al., 2023)) at project outset and iteratively guides fidelity to participatory goals.

7. Open Problems and Future Research

Persistent gaps and open questions include:

• Defining radical participation: Few studies articulate or operationalize criteria for “ownership” or “co-decision” as distinct from consultation or involvement (Delgado et al., 2023).
• Metrics for participatory impact: Quantitative evaluation of PD effectiveness—beyond process metrics to actual shifts in artifact function, power, or user outcomes—remains underdeveloped (Wacnik et al., 2024).
• Equity and intersectionality: Empirical and theoretical treatments of how participatory structures reproduce, challenge, or ameliorate inequities are needed, alongside formal metrics for participation depth and inclusion (QueerInAI et al., 2023).
• Scalability and generalization: Methodologies for scaling PD to large, diverse, or asynchronous populations without loss of engagement or agency are an active area of investigation, especially in open-source and crowd-sourcing contexts (Hellman et al., 2021, Cumbal et al., 23 Jun 2025).
• Integration with policy and governance: The use of IGAI and other boundary objects in civic, architectural, and municipal contexts raises open research questions on regulatory frameworks, long-term influence on built form, and sustainable stakeholder governance (Guridi et al., 2024).

Participatory Design Studies, in their diverse instantiations, continue to advance the co-creation of technological systems that are robust, trusted, contextually situated, and ethically responsive—yet the field is actively dialoguing with outstanding questions regarding depth, equity, impact assessment, and sustainable participation (Wacnik et al., 2024, Delgado et al., 2023, QueerInAI et al., 2023).