It makes a case from our own record at DreamSpace Academy, sets out the argument against it, and marks how well each claim is backed by evidence. For what we actually run, the main pages are the description to rely on. This essay gives the reasoning behind them.
Why our work reaches into Grade 6 rather than starting at the graduate
One of three essays on how our education work and the rest of DreamSpace Academy fit together.
In short. An organisation building an innovation and entrepreneurship ecosystem reaches down into Grade 6 rather than starting with graduates. The case below comes from our own experience at DreamSpace Academy, including the parts that argue against it.
Evidence marks. Accounts of what happened come from our own reports. Decisions come from our written plans. Marks: 🟢 backed by original documents or by more than one source; 🟡 supported only in part, or by one source; 🔴 stated, and still being tested.
The evidence for starting early is graded 🟡. It is real, and there is also published evidence against it, set out in section 7. The claim is that “the window argues for early”, never that “early is proven.”
1. Why we moved into schools
For our first seven years we worked with people aged 18 and over. Bootcamps, incubation, labs and the venture pathway were all aimed at young adults who had finished school.
Then we reorganised our work. Our concept report gives the reason in one sentence:
“Intervening at this age proved late for long-lasting transformation.” 🟢
We split our work into Empowerment Inside School and Empowerment Outside School. EIS, the in-school half, was something we had never tried before 🟢:
- it stays with the same pupils for six years, from Grade 6 to Grade 11;
- it is free to underserved government schools;
- it runs on the school’s own timetable;
- pupils learn by working on challenges rather than by memorising.
We concluded that we had been working at the wrong end of the problem. The conclusion came from our own results, which showed we were arriving too late. It did not come from a theory.
2. What the research says about starting early
We found the research afterwards, and the research agreed with what we had found. So the evidence was not picked to justify a decision already taken.
The skills behind starting a business grow in stages as children develop, and the first stages open earlier than most programmes start.
- Abstract business ideas wait on how a child’s thinking develops. Children cannot reliably tell the difference between owning capital and doing the work until about age 12 or 13 🟢. They can understand profit in middle childhood, but only once they have been taught to compare prices 🟢.
- Saving and selling can be learned from about age six, in a hands-on, playful setting 🟢.
- The most relevant randomised trial (BizWorld, about 2,751 children aged 11 to 12) shows clear gains in the personal qualities behind entrepreneurship: persistence, creativity and planning ahead 🟢.
- Only one study compares starting early with starting late. This randomised field experiment in the final grade of primary school found the same gains in mindset, and recorded that the same programme given to teenagers had no effect or a negative one (Huber, Sloof & Van Praag 2014). This finding is taken from a review that summarises the study.
The claim is narrow. It is not that children learn faster. An ecosystem needs people who are willing to try, to keep going and to see a problem as something they can fix. That willingness can be shaped at eleven and is much harder to shape at twenty-two. Our four-stage Changemaker Journey is built on this: Explorer, Maker and Innovator are stages of purpose before they are stages of skill.
3. School: the one place that reaches everyone
The research on child development says early. A second, separate line of evidence says in school. It is the stronger of the two, because it is about who can be reached rather than about how children think.
Where the barrier is cost and distance, a makerspace that people walk into does not reach the child. In Nuwara Eliya, 26.7% of children who left school did so because the household could not afford transport or learning materials, and 13.3% had no way of getting to school at all 🟢. In Mullaitivu the barriers are different: 33.3% had not passed O/L, and 11.1% had to work for the family 🟢.
Neither of those children ever arrives at a makerspace. Both are in a classroom until the day they leave. The school is the only place that reaches everyone, including the people least able to sign up for something themselves. That is why the in-school half of our work is the half that carries our promise to include everyone.
The state has also created an opening. Sri Lanka’s curriculum reform makes ICT and “Entrepreneurship & Financial Literacy” compulsory essential subjects in Grades 6 to 9. It phases in from 2026 and covers all four grades by 2029 🟢. But the schools are not equipped for it:
- only 19% of type 2 and type 3 schools have IT labs, against 72% of top-tier schools (type 2 and 3 are the smaller government schools that do not teach up to A/L);
- in the technology stream there are 35.3 students per teacher, against an average of 12.8 🟢;
- in the estate sector 1.5% of households own a computer and 92.6% of internet access is by smartphone alone 🟢.
When the state has promised something it cannot yet deliver, there is demand for help. That gives a place to an organisation that already has trainers, a curriculum and lab equipment in those districts: a partner that supports the state system and feeds into it.
The limit: whether that place can be made formal (an MoU, a provincial agreement, recognition as a teaching partner) is still an open question. What exists today is narrower:
- written Zonal Education Office approvals in three zones (Batticaloa, Kantale, Nuwara Eliya);
- a rule for choosing schools that picks those lacking the basics: no ICT teacher, no computer lab, low O/L ICT results, rural or marginalised 🟢.
4. Readiness: learners need the basics before advanced work
We learned the third reason from our own teaching.
Across many sessions, one of our experienced trainers found that the biggest limit on teaching is not equipment 🟢:
“The biggest issue is not the lack of facilities, laptops, internet, or tools… The biggest issue is the mismatch between the expected level of the session and the actual readiness of the participants.”
“In many sessions, less than 10% of the students were actually fit for the planned content.”
A separate check on about fourteen students chosen by interview found that five could copy a Scratch password-checker after seeing it demonstrated, and none could build one on their own. The four who finished fastest could not explain why the skills mattered 🟢. Finishing a task is not the same as being able to use the skill. Attendance drops off at the point where the work gets harder.
The figures for the region point the same way: literacy in the estate sector is 81.3% and falling, and 92.6% of internet access there is by smartphone alone.
A learner who has never used a desktop computer is not a learner who “just needs the advanced module.” Adding more classes on top of a starting level nobody has measured leads to the same drop-off. The answer is to measure where learners start, teach the basics first, and give enough hours.
All three are easier to give a child than an adult. A six-year programme in school has the starting test, the order of teaching and the hours built in. A four-week bootcamp for adults has none of them. The evidence on hours is clear: one workshop produced no significant gains where a programme of several sessions did 🟢.
Our analysis names the cheapest fix with the biggest return: test readiness with a short task when people join, instead of asking them whether they know the topic. It costs one task before the course starts.
5. Supply: children as the ecosystem’s future staff
The first three reasons are about learning. The fourth is about having enough people, the subject of the first essay in this series, Why an Ecosystem Builder Has to Teach.
Our hardest problem was never labs, funding or curriculum. It was identifying changemakers, because the old approach waited for people who were already motivated to come forward.
The fix was to build motivation through the Changemaker Journey rather than screen for people who arrive with it. That turns a search into teaching: instead of hoping to find people, we teach them. A school programme running from Grade 6 to Grade 11 across many schools is a teaching answer to that problem.
Earlier learners already show up in our own organisation:
- Alumni now lead much of our software, media and music work, and two are directors 🟢.
- When we took our approach to Erode, Tamil Nadu, it was “delivered by two people who are themselves products of the pathway” 🟢.
- The work reaches upward as well as downward. Our school work has grown from students to teachers, then to principals and zonal education staff, and most recently toward preschool teachers, through a partner that funds preschools in Sri Lanka 🟢.
There are two limits on this. First, those staff were all trained in Batticaloa, and Erode is our Batticaloa people travelling. It is not evidence that a new site can grow its own. Second, the one test of that failed: DreamSpace Mullaitivu closed. The cause was that nobody took over from the person running it: that person could not continue, no local successor with the interest was found, and the search looked in Mullaitivu only. Funding, day-to-day running and skills were not the cause.
The closure is also a reason to start early. If a site can keep going only while one person looks after it, it has to keep growing the next people to take over, and that has to start early.
6. What starting with children does not mean
It does not mean our education, or any MakerSpace, is for children. Wherever our programmes run, age does not decide where anyone is placed:
- it is not an entry rule;
- it is not an age band;
- it is not printed in publicity either, because a typical age printed on a brochure is read as a rule by exactly the people the programme is trying not to turn away.
Age is used only where safeguarding and duty of care require it. The STEAM programmes are for everyone. Placement is by evidence, through a placement check.
It does not mean entrepreneurship is taught to eight-year-olds. Business topics open up when a learner has the knowledge they build on, not at a set age. Early stages teach the entrepreneurial mindset through making, plus simple activities in selling and saving. Profit versus revenue, pricing and building a venture come in later stages, whenever a person reaches them.
It does not mean everyone is aimed at founding a company. The Changemaker stage opens onto an open field of outcomes, not a funnel. The routes we have seen so far (an impact venture, research, activism or community leadership, employment or freelancing, and Grow Your Own) are examples, not a boundary, and all are equally valid. The research warns that pushing students into starting businesses, when they would not otherwise have chosen to, can lower their outcomes 🟡, and we follow that warning.
It does not mean the school programme is a device donation. A laptop in a school with no ICT teacher does not get used. What the school gets is the machine plus a trained facilitator, a curriculum that follows the national syllabus, zonal approval, and a test at the start and the end.
It is not charity handed to a school. The clearest failure so far came from that misunderstanding. EIS was ended at Dickoya Tamil Maha Vidyalayam in 2026, and our progress report gives the reason in the school’s own terms: “They see us as an NGO and want us to buy things for the school” 🟢. When a school expects equipment rather than skills, the programme collapses, however much the students want it. In a region used to hand-outs, a free programme is read as a supply of things. Any programme of this kind has to plan for that.
7. Evidence against starting early
The case for starting early is graded 🟡, and it stays 🟡. Two published findings pull against it:
- The effects of personal qualities such as persistence on later life are mixed. A screen of 9,553 publications (Smithers et al. 2018, Nature Human Behaviour) found likely effects ranging from negative, through none, to positive, with signs that small studies and positive results were over-represented 🟡.
- Early gains can fade. A large school reform in Rwanda raised entrepreneurship after one year, but the gain had faded by year three, and employment was 6 percentage points lower. The losses fell mostly on students pushed into starting businesses who would not otherwise have chosen to 🟢.
The claim therefore reads:
The developmental window argues for starting early, and the one clean early-versus-late comparison in entrepreneurship education favours early. That is not the same as “early intervention is proven.”
Three more gaps apply to our own work:
- There is no evidence from Sri Lanka on the effects of teaching entrepreneurship at a young age, only government plans to do it. The best evidence from low- and middle-income countries comes from Rwanda, and nobody has tested whether it holds in post-conflict, multilingual, rural Sri Lanka 🟢.
- Sri Lanka does not appear in the maker-education research at all, across five separate reviews of the published studies 🟢.
- A deliberately sceptical search found no clear result linking rural makerspaces to income. That is why the Earn claim is a hypothesis being tested rather than a finding 🔴.
These gaps are also what we can offer a research funder. A region nobody has measured is a region nobody funds. We have eight years of work in one place that can be followed over time, a set of case studies, two peer-reviewed papers and a PhD study we have hosted. Few other organisations are in a position to produce this evidence.
8. The school programme so far
We keep two kinds of figure for the programme. Enrolment is who was signed up. Reach is who turned up. We never combine the two, and neither stands in for the other.
8a. Reach: who took part
These figures come from our programme reports:
| Cohort | Period | Reach |
|---|---|---|
| Batticaloa ICT pilot | July to December 2025 | 194 students in Grades 6 to 9, at 5 schools |
| Hatton, robotics and ICT | Late 2025 | 306 students at 2 schools; paused by the floods, resumed in January 2026 |
| Nuwara Eliya robotics | April and May 2026 | 114 students at 4 schools, 13 sessions, 21 projects |
| Batticaloa Grade 8 | January and February 2026 | 72 students at 5 schools |
8b. Where it runs
Batticaloa and Hatton since 2025, Kantale since 2026, and Nuwara Eliya and Kathiraveli from October 2026, through GROW and Kathira Greens. Mullaitivu is planned and waiting for zonal approval 🟢. We will publish the 2026 enrolment figures once they are final.
8c. Measuring learning
The programme requires a mid-term and a final test. We will publish the 2026 results once they are in.
Funding. EIS is free to schools and is paid for jointly by its partners, each covering part of the cost: IFNB (Germany), Micro-Tech CNC Pvt Ltd (India) under a signed agreement, and the Sabrina Yusoof Foundation 🟢. We have set out the least the programme needs each month to run each hub, covering staff, buildings and equipment, and materials. We do not publish those figures yet.
9. Summary of the argument
We decided to work with children after concluding, from our own results, that starting at eighteen was too late. The research on child development we found afterwards agreed, with cautions we have kept. Three more reasons support the decision:
- the school is the only place that reaches the child who cannot afford the bus fare;
- readiness, not equipment, is the biggest limit on every programme we run, and only a programme that stays in school for years has the starting test, the order of teaching and the hours to fix it;
- an ecosystem that cannot grow the next people to look after it closes when one person leaves, and we have already seen that happen once.
By our own reasoning, the children in a Grade 6 ICT class in Kantale are the ecosystem’s future staff. We cannot yet prove how many of them will become that.
A note on sources
The account of the school programme comes from our own documents: the concept report, the thirteen-phase written procedure, and our reports on locations, reach, funders and staffing. It also draws on the hubs’ monthly reports, the programme’s document checklist, the history of our funders, and case studies on learner readiness.
The written plans we drew on are our analysis of why young people struggle to earn, our description of who we are, the DreamSpace Lifecycle, the Changemaker Journey and the curriculum, and our decision to stop using age to place learners.
Statistics come from their publishers: the Department of Census and Statistics Labour Force Survey 2023 and Computer Literacy Bulletin 2024, UNDP and OPHI district profiles 2023, and the World Bank Sri Lanka Development Update (October 2024). Intervention evidence is cited to the published studies named in the text; the full list is on the references page.
Some figures often quoted on this subject have been shown to be wrong, and we use none of them.
Two common mistakes to avoid. The North and East do not have Sri Lanka’s worst youth unemployment: the Northern Province has the lowest, at 14.9%. The real problems are underemployment and informal work in the East, and people being trapped in the estate belt. And the multidimensional vulnerability index (2022 to 2023) and the multidimensional poverty index (2019) are different measures. They must never be presented as one figure changing over time.