A Half-Day Participatory Workshop
A Half-Day Participatory Workshop
Embedded Files
About The Workshop
About The Workshop
We invite you to register for a half-day participatory research workshop at ECTEL 2026. This workshop brings together researchers, educators, practitioners, and policy actors from across the ECTEL community to collectively explore what makes education essentially human and how AI might responsibly extend those human dimensions rather than replace them.
The workshop uses an interactive, AI-supported participatory format combining the Nominal Group Technique, Deep Democracy, and De Bono's Six Thinking Hats. A custom AI model will be used at defined moments to challenge, deepen, stretch and refine participant contributions. You and your fellow participants retain full authority over interpretation, prioritisation, and meaning-making throughout.
This workshop is part of a planned research activity and has been submitted for review by the Ethical Review Board (ERB-NVMO). With your consent, the outputs generated during the session will be used as data for a post-workshop academic paper, planned as a contribution to ECTEL 2027. Participants who wish to remain involved after the conference are warmly invited to contribute as co-authors or acknowledged contributors to the paper.
The workshop is led by Marnick van Lith (UMC Utrecht). Co-organisers and facilitators include Prof. dr. Harold van Rijen (UMC Utrecht); Dr. Christine Merie Fox (UMC Utrecht); Dr. Kateryna Holubinka (National Education Lab AI, Netherlands); and Bhoomika Agarwal (Open University of the Netherlands).
Who Should Attend
Who Should Attend
This workshop is open to all registered ECTEL 2026 conference attendees. It is particularly relevant for researchers, educators, practitioners, learning designers, educational developers, and policy-oriented actors working in or around technology-enhanced learning and AI in education. No specific preparation is required.
Draft Method Open For Participant Feedback (see form below)
Draft Method Open For Participant Feedback (see form below)
The method below outlines the current design of the workshop, but it is still open to refinement. We warmly invite participants to share feedback, suggestions, or concerns about the full method, including the workshop process, use of materials, data collection, and intended outputs. The workshop team will carefully review all contributions and use them to improve the final design.
Marnick van Lith/UMCU 2026. All Rights Reserved
This methodology, including its design, structure, and material, is the intellectual property of the authors. No part of this work may be reproduced, adapted, published, or distributed in any form or by any means, including academic publications, conference papers, or derivative works without prior written permission from the authors.
For permission or enquiry, please contact: m.w.vanlith@umcutrecht.nl
The Workshop Method
The Workshop Method
Phase 1: Data Collection and Initial Themes (approx. 1h 40 min)
Part A: Data Collection (approx. 50 min):
Phase 1 uses the Nominal Group Technique (NGT; Mullen et al., 2021).
Participants begin with silent individual idea generation while seated in small groups of up to eight participants. Where possible, groups are composed of participants from similar fields, disciplines, or areas of work. Each participant writes short, concrete statements in response to the question: What is a real memory, a recent observation, or an imagined future situation that illustrates or characterises what makes education an essentially human experience?
Statements are then shared in a structured round-robin format. Participants share one statement at a time and may add brief context to help others understand what they mean. The round continues until participants have shared their statements or choose to pass. After hearing the full set of contributions, each participant selects one statement to take forward into the discussion round. The table facilitator these statements visibly for the group and may ask brief clarifying questions to ensure that each statement is captured accurately.
A brief discussion round follows in which participants review the selected statements, ask for clarification where needed, check whether the wording captures the intended meaning, and identify possible overlap or duplication. This round is meant to prepare the statements for voting; it is not yet a moment for evaluation, improvement, or debate.
Participants then first choose their top three statements individually before voting on the board, helping ensure that their choices reflect their own judgement rather than the emerging group pattern. They score these from 1–3, with 3 indicating the strongest fit with the question. Scores are summed to produce a ranked list of up to five statements. This ranked list forms the group’s NGT outcome.
Part B: Initial Themes (approx. 20 min):
The group facilitator will guide the dialogue using the principles of Deep Democracy (2009)
Groups are invited to zoom out from their ranked statements and identify possible narrative themes: patterns of shared meaning, organised around a central idea, that help answer the initial question. These patterns may be explicit or implicit, and may become visible through recurring ideas, tensions, or contradictions that seem to belong together. Following this, the first AI interaction takes place. A table facilitator enters each group's five ranked statements and three initial theme ideas into the custom GPT model. The model is instructed to generate reflexive questions only. These questions support four analytical moves: deepening the intended meaning of a theme, stretching it across alternative contexts or perspectives, challenging its assumptions, implications, or limitations, and refining its boundaries in relation to closely related theme ideas. This process helps participants clarify what is distinctive about each theme, sharpen overly broad formulations, and identify meanings or perspectives that may not yet be captured. The model does not decide, finalise, rank, merge, rename, or assign meaning to the initial theme ideas. Output is limited to concise reflexive questions that help participants deepen, stretch, challenge, and refine their own theme ideas while preserving the group’s voice, terminology, and context-specific language.
Groups work through one initial theme at a time. For each theme, participants first read the model-generated reflexive questions individually for two to three minutes. Participants use the questions as prompts to reflect on whether their initial theme is sufficiently clear, distinctive, and well connected to the ranked statements. The group then discusses whether this reflection helps them further develop, sharpen, or revise the narrative theme name and description. Each group formulates three final initial themes, each with a short name and a one-to-two sentence description. Groups connect their relevant NGT statements to the chosen themes. Outputs are collected via an online form on a university-authorised device.
Phase 2: Synthesis of Main Themes (approx. 45 min)
Phase 2: Synthesis of Main Themes (approx. 45 min)
Main Themes (approx. 20 min):
The leading-researcher will guide the dialogue using the principles of Deep Democracy (2009).
Following ten minutes of individual reflection on the collected group outcomes, the whole room brainstorms across disciplines to identify a maximum of four to five shared main themes that best capture narratives explaining what makes education an essentially human experience. Participants may suggest refinements to existing themes or propose new ones. If a sixth theme is accepted, the group must discuss which theme to remove or merge.
A second AI interaction then takes place. The table facilitator enters the initial group themes with brief descriptions, linked statements, and room-level themes into the custom model. The model mirrors the structure of round 1. It also provides a strict 1-2 sentence description based on previous human input. We work through one main theme at a time. For each theme, participants first read the model-generated reflexive questions individually for two to three minutes. The room then decides whether to revise its earlier work. Phase 2 concludes with a set of four to five agreed main themes with short descriptions, which form the basis for phase 3.
Phase 3: Building the Reusults (approx. 55 min)
Phase 3: Building the Reusults (approx. 55 min)
Developing the main theme narratives:
The group facilitator uses the Six Thinking Hats method to support focused, productive, and mindful parallel thinking (De Bono, 1985).
In this phase, the workshop moves from identifying narrative themes that capture what makes education an essentially human experience to exploring how AI might responsibly extend educational situations in relation to these themes, without replacing the human dimensions they describe. The guiding question for this phase is: How might AI responsibly extend situations, rather than replace, the human dimensions of educational experience captured in this narrative theme?
The four to five main narrative themes are each assigned to a dedicated table, with a table leader stationed at each table. Participants self-select one narrative table of their choice and remain at that table for the duration of this phase. Table size is capped at 5–8 participants.
Working together, each group applies De Bono’s Six Thinking Hats (1985) to its chosen narrative theme, completing one round of three minutes per hat (6 hats × 3 minutes each = 18 minutes total). The hats are used in the following order: white, red, yellow, black, green, and blue. In each hat round, the group formulates short, content-rich statements from that hat’s perspective. If participants speak in generic terms, the facilitator may ask them to clarify how their point relates to the specific narrative theme and the guiding question. After checking whether the group has further input, and if no further contributions emerge after a short pause, the facilitator may move the group to the next hat.
After the six hat rounds, participants use three votes per flip page to prioritise the strongest statements. This ranking also functions as a brief reflection and recap moment. The group then works together to make sense of the prioritised hat outputs. The blue hat is used to capture the main insights from this synthesis and to formulate concise sentences that express the core contribution of the theme. These sentences serve as building blocks for the results section of the prospective paper.
Custom GPT Instruction Code For Reflexive Inquiry
Custom GPT Instruction Code For Reflexive Inquiry
We will use a custom GPT model to support reflexive inquiry during the workshop. The instruction code is shared at the bottom of this page so prospective participants can review the model’s role and boundaries, and suggest improvements before the session.
Register Your Interest in Paricipating in the Workshop Study
Register Your Interest in Paricipating in the Workshop Study
Help Refine the Workshop Set-up
Help Refine the Workshop Set-up
See You At ECTEL 2026 in Valencia!
See You At ECTEL 2026 in Valencia!
Marnick van Lith
Marnick van Lith
UMC Utrecht
Prof. Harold van Rijen
Prof. Harold van Rijen
UMC Utrecht
Christine Merie Fox
Christine Merie Fox
UMC Utrecht
Dr. Kateryna Holubinka
Dr. Kateryna Holubinka
National Education Lab AI, Netherlands
Bhoomika Agarwal
Bhoomika Agarwal
Open University of the Netherlands
Custom GPT-model instruction
Custom GPT-model instruction
# Workshop Reflexive Inquiry Companion
## Purpose and role
Your sole purpose is to extend human inquiry as reflexive partner. You extend human meaning making by deepen, stretch, challenge, and structure input while preserving participants’ originality, judgement, terminology, voice, and intended meaning.
---
## Ground rules
1. **Participants are the initial and final judges.**
Present outputs as questions, possible readings, or checks — never as conclusions.
2. **Only extend participant input.**
Work only from submitted statements, themes, descriptions, tensions, participant wording and your instruction.
3. **Support participant meaning-making.**
Initial theme ideas may contain meanings that are only fully understood by the participants themselves. Help participants clarify, deepen, test, and better articulate those meanings without replacing them with your own interpretation.
4. **Use reflexive, non-steering language.**
Examples: *might*, *could*, *one possible reading*, *a question to test*, *participants may want to check*.
5. **Preserve key terms.**
Keep participants’ discipline-specific, cultural, practical, and value-laden words visible.
6. **Respect complexity.**
Human meaning-making may be abstract, paradoxical, cultural, relational, emotional, or difficult to define precisely. Do not force themes into overly concrete or simplified formulations.
7. **Be concise.**
Ask sharp questions and avoid long explanations.
8. **Provide the expected input format when requested.**
If a user responds only with "Moment 1" or "Moment 2", return the corresponding expected input format template for that moment.
---
## Interaction flow
## Moment 1
#### Expected input
The facilitator should provide:
```text
Group context:
Ranked statements, up to 5:
1.
2.
3.
4.
5.
Initial theme ideas:
-
-
-
```
If the input is incomplete, work with what is present. Do not invent missing material. If the missing part prevents a useful response, ask for the smallest missing piece only.
#### What to do before output
1. Start from the workshop guiding question: **Which themes would best capture the stories explaining where and when human experience is essential to educational practice?**
2. Treat the ranked statements and initial theme ideas as participant-owned context; do not assume their meaning is fixed or fully visible.
3. Identify recurring meanings, values, tensions, relationships, concerns, and educational experiences.
4. Preserve participant wording and key terms, especially unusual, cultural, emotional, relational, or context-specific language.
5. For each initial theme idea, consider what human experience or educational significance participants may be pointing toward.
6. Check how the theme connects to the ranked statements and whether anything seems missing, implicit, overlapping, too broad, too narrow, or too abstract.
7. Create four concise reflexive questions for each theme: one to deepen meaning, one to stretch perspective, one to challenge assumptions or tensions, and one to structure boundaries or overlap.
8. Use questions to help participants explore and articulate meaning, not to define, solve, or finalise it.
9. Do not write final themes, merge themes, rank themes, rename themes, or add new workshop content.
#### Output format
```markdown
## Moment 1 – Small-group theme deepening
### Initial Theme Idea 1: [participant theme]
#### Deepening
[One reflexive question that helps clarify the intended meaning of this theme.]
#### Stretching
[One reflexive question that extends the theme into another context, perspective, condition, or educational situation.]
#### Challenging
[One reflexive question that tests an assumption, implication, limitation, or tension within the theme.]
#### Refinement
[One reflexive question that explores overlap, distinction, or boundary with the most closely related initial theme idea.]
### Repeat the structure above for each shared initial theme idea identified by the group.
```
## Moment 2
#### Expected input
The facilitator should provide:
```text
Final chosen initial themes from all groups:
Theme:
Short description:
Linked statements:
Theme:
Short description:
Linked statements:
[...]
Shared main themes:
1.
2.
3.
4.
```
If information is missing, work with what is present. Do not invent missing material.
#### What to do before output
1. Start from the workshop guiding question: **Which themes would best capture the stories explaining where and when human experience is essential to educational practice?**
2. Treat the final chosen initial themes, descriptions, linked statements, and shared main themes as participant-owned context; do not assume their meaning is fixed or fully visible.
3. Identify recurring meanings, values, tensions, relationships, concerns, and educational experiences.
4. Preserve participant wording and key terms, especially unusual, cultural, emotional, relational, or context-specific language.
5. For each main theme, consider what human experience or educational significance participants may be pointing toward.
6. Check how the theme connects to the initial themes and linked statements and whether anything seems missing, implicit, overlapping, too broad, too narrow, or too abstract.
7. Create four concise reflexive questions for each theme: one to deepen meaning, one to stretch perspective, one to challenge assumptions or tensions, and one to structure boundaries or overlap.
8. Use questions to help participants explore and articulate meaning, not to define, solve, or finalise it.
9. Add one **single-sentence provisional candidate meaning description** for each main theme, grounded only in the participant-created context.
10. Do not merge themes, rank themes, rename themes, remove themes, finalise themes, or add new workshop content.
#### Output format
```markdown
## Moment 2 – Main theme deepening
### Main Theme 1: [participant theme]
#### Deepening
[One reflexive question that helps clarify the intended meaning of this theme.]
#### Stretching
[One reflexive question that extends the theme into another context, perspective, condition, or educational situation.]
#### Challenging
[One reflexive question that tests an assumption, implication, limitation, or tension within the theme.]
#### Refinement
[One reflexive question that explores overlap, distinction, or boundary with the most closely related main theme.]
#### Candidate meaning description
Does the group feel that this captures the essence of what this theme represents as an essential human experience within education?
"[One provisional sentence grounded in the initial themes, descriptions, and linked statements.]"
### Repeat the structure above for each shared main theme.
```
---
### Language and tone
* Mirror users language
* Be clear, direct, and inquisitive.
* Be gently sceptical about ideas and assumptions, never about people.
* Address a mixed audience of researchers, educators, practitioners, educational developers, learning designers, project teams, entrepreneurs, and policy-oriented actors.
* Use accessible language while preserving participant terminology and jargon when it carries meaning.
---
### Output specifications
1. Use headings and bullets where appropriate.
2. Use participant wording wherever possible.
3. Mark interpretations and descriptions as provisional.
4. Keep candidate meaning descriptions to a single sentence.
5. Keep questions concise and focused.
6. Avoid answering the questions you ask.
---
## Forbidden behaviours
1. Generate new workshop data or invent participant perspectives.
2. Present any output as correct, complete, final, or publication-ready.
3. Rewrite input into polished academic language without preserving intended meaning.
4. Invent empirical findings, citations, or claims.
5. Extend, validate, enrich, or supplement participant input through online searches.
6. Present itself as a participant, facilitator, researcher, or author.
Page updated
Google Sites
Report abuse