What happens when you try to run a nationwide Wikipedia quiz competition for secondary school students — in a country where most teens have never edited (or even seriously read) Wikipedia — with a tiny team and a volunteer network that turned out to be far thinner than expected?
That’s essentially what Wikimedia Bangladesh attempted in early 2026, with WikiScholar. And while it wasn’t perfect, it taught us more in four months than we could have planned for in a year.
The idea behind WikiScholar
The premise was simple: take 12 carefully selected Wikipedia articles, ask secondary school students to read them, then test their understanding in a multi-round quiz. The hope was that in doing so, students who had never visited Wikipedia for anything beyond copy-pasting homework would start to see it as something more — a living, collaborative body of knowledge that they could eventually contribute to themselves.
WikiScholar was also designed around a vision of who gets to participate in the free knowledge movement. Too often, Wikimedia outreach in Bangladesh reaches university students and adults. We wanted to go younger. Secondary school students — grades 6 through 10 — are at exactly the age when reading habits and digital curiosity form. If we could reach them then, the long-term impact could be significant.
So we launched. For the first time. In late 2025.
How the competition worked
We divided participants into two groups: Group A (grades 6–8) and Group B (grades 9–10, including SSC candidates). The structure had two main phases.
The first phase was a 30-minute online MCQ test — 60 questions — hosted on Google Forms. Top performers from each school would qualify for an offline second round, and from there, the top three per group per school would advance to the regional event.
The second phase was the regional competition, held in two cities: Dhaka (at Daffodil Plaza) and Rajshahi (at Rajshahi University‘s Dean’s Complex). These were in-person events with 100 questions in 50 minutes, prize giving, and a real celebration of the students who made it through.
A national round was planned for the future — for 2026’s inaugural edition, we kept the scope regional.
The numbers: what actually happened
WikiScholar 2026 was never just a quiz — it was, at its core, a Wikipedia awareness campaign. And by that measure, the reach was significant: across the participating schools in Dhaka, Rajshahi, and Chandpur, we connected with an estimated 10,000+ students through awareness sessions and outreach activities. For many of them, it was the first time someone had walked into their classroom and explained what Wikipedia actually is, how it works, and why it matters.
Of those, 1,499 students took the next step and registered for the competition — from 21 schools across the three regions. Roughly 70% came from Dhaka, 30% from Rajshahi. The split between groups was fairly balanced — 56% in Group A, 44% in Group B. For a first-ever edition of a brand new competition with no prior brand recognition, that’s a meaningful base to build on.
One number we’re not satisfied with: only 24.6% of registrants were female students. That gap didn’t escape us, and it’s something we’ve flagged explicitly for next year.
From the registered pool, 334 students sat the online round — 156 in Group A, 178 in Group B. In Rajshahi, no school in either group reached the threshold of 30 participants we’d set for unlocking a second offline round, so we adapted: the top three from the online round in each group went straight to the regional event. Some Dhaka schools managed second-round offline tests, and the three Chandpur schools came together to hold a joint second round. The regional finals in Dhaka and Rajshahi brought the competition to a proper close — with prizes, a ceremony, and students who had genuinely worked hard to get there.
Some Dhaka schools did manage second-round offline tests. The three Chandpur schools even came together to hold a joint second round. But across the board, the attrition between registration and actual participation was the biggest surprise of this first edition.
The challenges — and why honesty matters here
We kept detailed notes on what went wrong, because pretending things went smoothly doesn’t help anyone planning the second edition.
Technology was the first headache. We wanted a proper proctored online exam — no tab switching, no copy-paste, AI detection. We found platforms that did all of this. They were also prohibitively expensive. So we used Google Forms. It worked in the sense that it delivered the questions and collected answers. It didn’t work in the sense that we found clear evidence of some participants using AI to answer questions. That’s a solvable problem, but it requires investment we didn’t have this time.
Volunteer support was thinner than expected. WikiScholar was designed as a volunteer-driven program. In Dhaka and Rajshahi, some volunteers stepped up and were genuinely invaluable. But across the broader network, the expected coordination didn’t materialize.
Part of this is a structural reality that most Wikimedia organizers quietly know but rarely say out loud: Wikipedia attracts a particular kind of contributor. The people who spend hours improving articles, fixing citations, and debating policies on talk pages are often not the same people who want to show up at a school, introduce themselves to a teacher they’ve never met, and convince thirty teenagers to register for an online quiz. That’s a different skill set, a different energy, and honestly — a different personality. Wikipedia’s contributor culture skews strongly toward introverted, screen-side work. Asking those same volunteers to do legwork and in-person outreach is asking them to step well outside their comfort zone, and many simply won’t.
When your model depends on volunteers to mobilize schools, coordinate logistics, and manage on-the-ground communications — and the volunteers aren’t there at that scale — everything collapses onto a very small core team. This is something we’re actively rethinking. One direction we’re seriously considering is looking beyond the Wikimedia movement itself for volunteer support. Movement volunteers are brilliant at what they do, but fieldwork isn’t always what they signed up for. For an education-focused program like WikiScholar, it may make more sense to recruit volunteers directly from the participating schools — students, teachers, or alumni who are already embedded in those communities, already trusted by the administration, and naturally motivated to see their school do well in the competition. A student ambassador at Ispahani Cantonment Public School is going to get thirty classmates to register far more effectively than a Wikipedia editor who has never set foot there.
Communication broke down in places. We used Mail merge for emails and a bulk SMS service for texts. About 8% of emails bounced. A significant number of SMS messages never arrived despite valid numbers — and the provider’s customer support had no resolution. These are mundane logistical problems, but in an outreach program targeting school students, they translate directly to no-shows.
The part that made it worth it: student feedback
After the regional events wrapped up, we sent a feedback form to participants. Fifty-eight students responded in just seven days.
The responses were genuinely moving to read.
Before WikiScholar, 34.5% of respondents said they had rarely or never used Wikipedia. After the competition, 87.9% said they now view Wikipedia positively — either as an important knowledge resource or as a good medium for learning. That’s not a trivial shift. These are students who spent weeks reading Wikipedia articles to prepare and came out the other side with a different relationship to the platform.
91.4% rated the competition as “good” or “very good.” 94.8% said they would definitely participate again. And 98.3% said they want to keep receiving updates from us by email.
When we asked for open-ended suggestions, the students’ feedback was thoughtful and specific: expand to all districts; add analytical questions (not just fact recall); publish everyone’s scores so results feel transparent; bring in preparatory classes or study sessions before the exam; and please — do something about the AI cheating problem.
One thing the students kept asking for: take this nationwide.
What comes next
We’ve learned enough from 2026 to propose a substantially redesigned model for 2027.
The current approach — visiting every school individually to build participation — is not sustainable at our capacity. For the next edition, we’re proposing a more scalable marketing strategy: CentralNotice banners on Wikipedia itself (geo-targeted by division), formal letters to top schools across all divisions, Facebook and Google Ads, school ambassador programs with certificates and souvenirs as incentives, and media partnerships to live-stream regional events.
On the technology side, we want to build our own web application — one that handles registration, exam delivery, communication, and results end-to-end. The platform would auto-slot participants into exam windows based on server capacity, pull questions automatically from a question bank, and include basic proctoring (copy-paste disabled, auto-close on tab switch). This would solve several of our worst problems at once.
We’re also rethinking the competition structure. The proposal is to move all offline exams to the divisional level, invite the top performers from each upazila/district to compete there, and hold a national final in Dhaka for the best from across the country.
And on gender parity: we want to proactively visit girls’ schools in major cities, because passively waiting for female students to register clearly isn’t enough.
A note on what this is really about
WikiScholar isn’t just a quiz. It’s an attempt to change who feels like they belong in the free knowledge ecosystem — to reach students who’ve never thought of themselves as potential Wikipedia readers or editors and plant a seed.
The ~60 feedback responses we received are a small sample. But when a student who had never visited Wikipedia before the competition tells you they now see it as an important knowledge resource and want to come back next year, that’s not a small thing. That’s exactly what Wikimedia outreach is supposed to do.
We’ll be back in 2027. With a better platform, wider reach, and hopefully a few more volunteers.
Can you help us translate this article?
In order for this article to reach as many people as possible we would like your help. Can you translate this article to get the message out?