Situated Participatory Design

• Situated Participatory Design is a method that embeds co-design activities in participants’ everyday contexts, emphasizing agency development before participation.
• It has been operationalized across various domains, including political empowerment, healthcare, human–robot interaction, and AI alignment using iterative, context-specific protocols.
• By anchoring design in real-world settings with multimodal tools, sPD redistributes power and reveals practical constraints, leading to ethically transformative outcomes.

Situated Participatory Design (sPD) is an evolution of traditional Participatory Design (PD) that foregrounds the embedding of co-design activities within the everyday, contextual realities of participants. sPD is defined by two central commitments: (1) agency is cultivated prior to, and as a prerequisite for, meaningful participation, and (2) design activities are anchored in localized situations—be these physical, social, organizational, or epistemic—ensuring outcomes are not just theoretically inclusive but practically transformative. sPD has been operationalized in domains ranging from political empowerment of vulnerable populations and co-design of complex socio-technical infrastructure to the in situ refinement of human–robot interactions and the dynamic negotiation of AI alignment values (Gautam et al., 2020, Robert et al., 26 Mar 2025, Stegner et al., 2023, Arzberger et al., 22 Jan 2026).

1. Theoretical Principles: Agency, Situatedness, and Participation

sPD re-conceptualizes two key elements of design:

Agency before Participation: In contrast to classical PD, which assumes participants possess the inherent agency to envision and enact future scenarios, sPD specifically addresses contexts of vulnerability or power asymmetry. Design encounters are sequenced so that participants first develop a sense of themselves as actors within their immediate environment (“agentic-future-envisionment”), before being asked to deliberate about or intervene in broader, institutional structures (Gautam et al., 2020). In mathematical notation:

• $A_s$ = agency in small-scale context;
• $P_s$ = participation in small-scale context;
• $A_L$ = agency in large-scale/institutional context;
• $P_L$ = participation in large-scale context.

sPD formalizes:

1. $\Delta A_s > 0 \implies \Delta P_s > 0$
2. $P_s \rightarrow \Delta A_L$
3. $\Delta A_L > 0 \implies \Delta P_L > 0$

Situatedness: Drawing from ethnographic and feminist theory, sPD insists that values, needs, and capacities materialize only in context and cannot be abstracted meaningfully from lived practices. The design process, therefore, is inseparable from the material, social, and organizational settings in which it is enacted (Robert et al., 26 Mar 2025, Arzberger et al., 22 Jan 2026).

Participatory Co-Construction: sPD positions users as epistemic agents who co-create design and alignment practices through hands-on, contextually embedded activities. This participation extends beyond consultation to ongoing, run-time interventions and feedback (Arzberger et al., 22 Jan 2026).

2. Domain-Specific Methodologies and Protocols

sPD is not a single methodology but an adaptable meta-framework operationalized through situated protocols, detailed in recent case studies:

Empowerment in Vulnerable Populations: Workshops for Nepalese sex-trafficking survivors employed low-literacy, multimodal artifacts (drawings, collage, oral narration), future-vision worksheets across six life domains, and problem-mapping posters. Activities were conducted in the shelter environment, leveraging everyday metaphors (e.g., family or village relations) to scaffold agency-building prior to institutional engagement (Gautam et al., 2020).

EHR Component Design in Healthcare: The ERIOS lab at CHU de Montpellier embedded sPD in the workplace, structuring interaction through three phases—Ethnographic Observation (Telling), Individual/Collective Prototyping (Making), and Realistic Simulation (Enacting). Boundary objects (paper and digital prototypes), iterative simulation with clinical artifacts, and a co-facilitation charter redistributed agency and surfaced tacit expertise (Robert et al., 26 Mar 2025).

Human–Robot Interaction with Older Adults: sPD for assistive robots in senior living facilities unfolded in three phases: (i) Discovery and User Enactment (real robot demonstration and scenario co-design in residents’ apartments), (ii) Simulated deployment (unsupervised, in situ testing), and (iii) Stakeholder reflection involving caregivers and organizational leadership. Tooling included Wizard-of-Oz teleoperation, multi-modal logs, and iterative scenario worksheets (Stegner et al., 2023).

AI Alignment as Situated Practice: Participatory workshops with LLM users, using two-week misalignment diaries, collaborative reflection on value conflicts, and generative interface prototyping. The process surfaced concrete user “roles” (adjuster, interpreter, non-engager, steward), translated metaphoric concepts into speculative interface proposals, and articulated design principles for real-time, context-responsive alignment management (Arzberger et al., 22 Jan 2026).

3. Conceptual Models and Frameworks

Relational Space Models: sPD projects often formalize the redistribution of agency using relational indices and process frameworks. For example, agency-shift can be represented as:

$$p_s^\text{final} = p_s^\text{initial} + \Delta p \ p_{\text{IT}}^\text{final} = p_{\text{IT}}^\text{initial} - \Delta p$$

Where $p_s$ and $p_{\text{IT}}$ are the agency indices for staff and IT specialists, respectively, and $P_s$0 is the shift induced by co-design interventions (Robert et al., 26 Mar 2025).

Process Frameworks: The Telling–Making–Enacting (TME) model organizes sPD workflows:

• Telling: Elicitation of lived narrative and pain points.
• Making: Collective prototyping and critique.
• Enacting: Realistic scenario simulation, often with disruptive real-world events.

Interactional Alignment Loop: In AI contexts, alignment is structured as an interactive sequence:

$P_s$1

This loop, in conjunction with interface affordances (sliders, prompt libraries, “abort” controls), constitutes a run-time instantiation of sPD (Arzberger et al., 22 Jan 2026).

4. Insights, Benefits, and Limitations Across Domains

Empirical Insights:

• Agency accrues through participation in small-scale, personally meaningful acts, subsequently catalyzing readiness for larger-scale or institutional action (Gautam et al., 2020).
• In situ, experience-based engagement reveals design breakdowns and emergent needs not available through simulated or abstract prototyping (Stegner et al., 2023).
• Dedicated relational spaces—physical (workshops, labs), material (props, dashboards), and digital (intranet forums, prompt libraries)—enable more equitable negotiation across stakeholder groups (Robert et al., 26 Mar 2025).
• Participatory value alignment in AI surfaces not only technical misalignments (hallucinations, overconfidence) but also social and ethical tensions, such as autonomy, interpretability, and collective responsibility (Arzberger et al., 22 Jan 2026).

Benefits:

• Enhanced user agency and self-efficacy, especially critical with marginalized or dependent populations (Gautam et al., 2020).
• Early, context-specific surfacing of practical and social constraints, facilitating robust design iteration (Stegner et al., 2023).
• Redistribution of power and expertise among participating groups, averting designer-centric or technologically deterministic outcomes (Robert et al., 26 Mar 2025).
• Creation of generative, reflexive interfaces that accommodate situated, evolving values rather than presupposing static, abstract value sets (Arzberger et al., 22 Jan 2026).

Limitations:

• Tooling and session design are resource-intensive, often restricting the scope of participant involvement per cycle (Stegner et al., 2023).
• Outcomes are often qualitative and emergent rather than quantitatively benchmarked, though future work may integrate metrics such as iteration counts or satisfaction ratings (Stegner et al., 2023).
• Longitudinal effects of agency-building and situated alignment require sustained engagement, often difficult under institutional or funding constraints (Gautam et al., 2020, Arzberger et al., 22 Jan 2026).

5. Design Challenges and Operational Considerations

sPD foregrounds key operational challenges:

• Negotiating Power and Authority: Activities must support transparent navigation of outsider/insider dynamics and organizational hierarchies, as exemplified by researcher–participant negotiations in vulnerable populations and clinician–IT relationships in healthcare settings (Gautam et al., 2020, Robert et al., 26 Mar 2025).
• Multimodal Artifacts and Accessibility: Low-barrier artifacts and non-textual modes (e.g., drawing, enactment, speech) are needed when working with participants with limited formal education or physical/cognitive constraints (Gautam et al., 2020, Stegner et al., 2023).
• Iterative, Flexible Tooling: Reconfigurable spaces, material “boundary objects,” and digital supports allow sPD to surface authentic input, facilitate negotiation, and support real-time iteration (Robert et al., 26 Mar 2025).
• Emotional and Ethical Safety: Representation of sensitive life events or trauma requires culturally resonant metaphors and opt-out mechanisms to ensure emotional safety (Gautam et al., 2020).
• Collective and Reflexive Tools: In AI applications, tools to capture misalignments, surface positionality, and foster communal reflection (dashboards, prompt libraries) are essential for collective sense-making and value co-construction (Arzberger et al., 22 Jan 2026).

6. Cross-Domain Generalizability and Future Directions

sPD demonstrates significant promise across a diversity of sociotechnical contexts:

• Transferability to new domains (assistive tech, public services, education, complex AI systems) is contingent on the careful adaptation of participatory protocols, context-specific artefacts, and stakeholder configurations (Stegner et al., 2023, Arzberger et al., 22 Jan 2026).
• Future research directions include the automation of iterative interaction pipelines, quantitative benchmarking of participation and satisfaction, scaling up to larger user populations, and the formalization of models mediating authority—particularly in the context of supervisory control and institutional oversight (Stegner et al., 2023).
• sPD suggests a reorientation of design expertise: from producing deliverables to enabling ongoing, collective, and context-resonant platforms for action across technical, organizational, and political registers (Robert et al., 26 Mar 2025, Arzberger et al., 22 Jan 2026).

7. Summary Table: Key Characteristics of sPD Instantiations

Domain/Application	Contextual Anchoring	Distinctive sPD Feature
Political empowerment	Shelter, village, and family relationships	Sequenced agency-building, multimodal
EHR co-design	Hospital-adjacent lab, clinical workflows	Telling-Making-Enacting, relational spaces
Human–robot interaction	Resident apartments, senior living facility	In-situ enactment, Wizard-of-Oz, reflection
AI value alignment	User work practices, LLM-based research	Misalignment diaries, participatory interface prototyping

Each instantiation adapts sPD principles—agency before participation, situatedness, participatory co-construction—to the material and sociopolitical exigencies of its context, leveraging context-specific protocols, artifacts, and iterative cycles (Gautam et al., 2020, Robert et al., 26 Mar 2025, Stegner et al., 2023, Arzberger et al., 22 Jan 2026).