The brief
A cohort that didn't choose this, doing work that can't afford to go wrong
NTU's Care Serve Learn programme places undergraduates into 40-hour service projects with young people, some from schools, some from homes, including youth with special needs and trauma histories. Up to 400 students would pass through a single day of training before starting that year, on a pass/fail basis, with most of them there because the programme required it, not because they chose it.
Engagement
How do you get people to care who didn't ask to be here?
Delivery
How do you make it work without expert facilitators?
Responsibility
Students go on to work with young people who may be vulnerable.
The curriculum had to create engagement, work safely in inexperienced hands, and build in a safety net, all at the same time.
What was built
One architecture, worn as three modules
Students choose two of three modules per session. Underneath, all three share the same shape: self, then relationship, then action. Every module asks students to connect an internal idea to something observable, a choice, a relationship, a step they can actually take.
Values
What matters to me?
Emotional Bank Account
How do I affect others?
Dreams & Aspirations
What will I do with it?
A three-tier safety protocol runs beneath every round, and a teacher or support worker is a non-negotiable presence at every real beneficiary session that follows. None of this arrived fully formed: two rounds of student testing and six structured reviews came before a single NTU student saw it on 5 September. That build process is its own case study. Read Part 1: Design and Build →
Delivery day
What actually happened, in the room, on 5 September
The training ran as two sessions, a morning and an afternoon, both youth cohorts drawn from the NTU student population this run was designed for. Every board, physical and digital, was live at once, which is the first time the full design had been tested outside a rehearsal.
The morning carried most of the day's volume, but the smaller afternoon cohort completed its groups at a comparable rate, evidence the design didn't depend on a particular room size or time of day to hold together.
What the boards showed
The numbers say participation held up. The boards show what it actually looked like
Every board reference below is paraphrased and unattributed, no names, no images, consistent with there being no photo or data consent collected on the day. On the shared poster where students drew what they wanted their future to hold, the same handful of themes kept reappearing across different students' sections: family, career or business ambition, travel, higher education, financial security. That matches the digital log almost exactly, where Adventure and travel and Meaning and purpose were the two most logged dream categories of the day. Two independent records of the same activity, a hand-drawn poster and a digital tally, landed on the same handful of priorities.
Values · gifting · talent · skills
With them.
Not to them.
The same phrase, close to word for word, on five or six separate boards from groups that had no way to see each other's work.
That's a stronger signal than any single student's board, because these groups had no way to coordinate. A value isn't something you simply possess, it becomes meaningful through what you do with it alongside other people. The Values module didn't have to announce that conclusion. The activity created the conditions for students to arrive at it themselves. What follows is a small sample of what individual groups did with the same prompts.
The dream sort
Independence, adventure, family, creative expression, four named categories, connected with a drawing of her own family at the centre.
Abstract prompt → concrete future.
The travel map
A whole sheet of landmarks, the Eiffel Tower, Big Ben, the Leaning Tower of Pisa, a parachutist crossing the page.
A purely visual answer to "where do you want to go."
Peer support, in the group's own words
Each person named their own driver, the group assigned roles, and one group wrote the literal line each role would say for the person they were supporting.
Encourager“You were put on this earth to do something.”
Doer“Book tickets for her, go with her.”
Idea Generator“Talk about it more, ask around.”
Individual driver → collective commitment.
The chain story
One member's imagined tech company funds a second member's Broadway musical, financed through a third member's business network. The whole group pictured itself in the audience.
Individual dreams → one shared picture.
The joke life plan
Drop out, become a CEO, fly to Zimbabwe, build a villa, find a wife, an exaggerated five-step plan mapped straight onto the module's four real categories: financial independence, travel, a stable home, family.
Humour still routes back to the real structure.
One small, human detail worth keeping without any personal or medical specifics: in one group, a student's stated driver, an active and healthy old age, traced back to watching a family member age well and stay active. A quiet anchor for an otherwise abstract category.
What the student facilitators learned
The real test wasn't the content. It was whether undergraduates could run it
45 responses were collected from the NTU students who went through the training. Here's what they said, in their own words, about what they learned about facilitation. They cluster tightly around five things: listening and empathy, ahead of everything else; psychological safety and inclusion, letting someone choose not to share, not saying "you should," making sure no one is left out; staying neutral rather than centring their own stories, helping "with" someone rather than "for" or "to" them; the craft of it, sharper follow-up questions, time management, reading a room's energy; and a fifth cluster, boundaries, that matters more than the rest.
Why this matters beyond delivery day
Nobody told these 45 students to think about a participant's life outside the room, or to name the programme's own safety framework. Several wrote, unprompted, that the young people they'll go on to work with may have complex relationships at home, and one invoked STOP, SLOW, MONITOR by name, the exact escalation language built into the training, without being asked to reproduce it. That's a stronger result than the module content landing. It means the facilitators picked up the posture the whole programme is built on, help with someone, not for them or to them, and the safeguard underneath it, on their own, straight out of running the sessions.
None of this reads like a manual being repeated back. It's the kind of language that shows up when someone has actually run a group, paid attention to what worked, and started worrying about the right things without being told to, which is exactly the outcome the training was designed to produce.
The numbers behind the day
Everything above, held up against the full board log
Every figure below is anonymous, drawn from the digital board event log, both sessions combined, and final.
Values carried the most volume by far across both sessions, with the Emotional Bank Account a solid second and Dreams the smallest, largely because Dreams only opened up properly once each session reached its second rotation: 4,016 recorded Values events, 2,266 in the Emotional Bank Account, 358 in Dreams. Activity followed the day's own shape rather than a steady pace: strong through both morning rotations, quiet over the lunch break, then a second rise across the afternoon's two rotations with a final burst just after 5pm as groups raced to finish.
Growth dominated the Values auction.
Values, Round 2: which values pulled the tokens
Growth leads for the day as a whole, with Courage and Purpose close behind. This order held in both sessions individually, so it's a real pattern, not an artifact of adding two sessions together.
Adventure and travel came first.
Dreams: categories and resource cards, full day
Adventure and travel leads the dream categories, with Meaning and purpose second. Freedom and independence overtakes Family and relationships for third place once both sessions are combined, a shift that wasn't visible from either session on its own. On resource cards, Planner edges out Connector, with Doer and Encourager tied for third.
Three relationship behaviours stayed contested all day.
Emotional Bank Account: ranked by how contested it was
All twelve actions, both sessions combined, ordered from most evenly split to most one-sided. Letting something go instead of raising it, checking in on someone every day, and offering practical help without being asked were the three that stayed contested all day, worth leading with in any debrief. Most of the rest read clearly as one or the other.
What this means
Three separate kinds of evidence, all pointing the same way
The original brief was to build something that would work for a large, partly involuntary cohort without leaning on psychological language or professional facilitation skill, and to do it safely in the hands of undergraduates. Delivery day produced three independent angles of evidence, and all three point the same direction.
Substance
303 students on Values rosters, specific and personal outputs
Delivery
45 facilitators adapting, observing, responding
Scale: participation held up across both sessions, over 300 logged entries in the Values module alone, 53 groups, thousands of board interactions across a morning and an afternoon.
Substance: where students engaged, the output was specific and personal, and in several cases the same idea surfaced independently across groups that never saw each other's work.
Safety: the facilitators' own account of what they learned shows the quality of delivery came from the structure the training gave them, not from expertise they didn't have going in.
What this confirms about designing for scale
Three working principles came out of the first live run, worth naming for anyone building a large-scale human skills programme of their own.
When hundreds of people move through an experience at once, engagement can't depend on facilitator charisma. The activity architecture has to create participation on its own.
The most compelling outputs weren't definitions of values or aspirations. They were travel maps, imagined companies, musicals, a joke life plan that still routed back to the real categories. The more personal the exercise, the more concrete the learning.
The facilitators didn't start as expert group leaders. The design gave them questions, structures, roles and time boundaries to work with, and they became more attentive by running it.
Notes on this case study
What's in scope, and what isn't yet
This case study is also, in a smaller way, a sample of the method behind it: design the conditions, make the learning observable, build for the person actually delivering it, then let the people in the room show you what the curriculum did. Start from a real constraint, test the design against people who'll tell you the truth, build in a way that's safe before it's polished, and measure what actually happened rather than what was hoped for. If you're building a large-scale training that has to work safely on the first real day it runs, this is the kind of work Human Fluency Lab does.