In October 2025 this was a club. A room, a whiteboard, and about a dozen people who wanted to do research and did not know how to start.
Ten months later it is more than 300 students across five provinces and two U.S. states. That number is not the interesting part. What the students did once they showed up is, and some of it was the opposite of what we expected.
Nobody was missing ability. They were missing a first step.
We assumed we would spend most of our time teaching skills. Statistics, reading papers, methods. That is not where students got stuck.
They got stuck before that, on a question so basic almost nobody says it out loud: am I allowed to do this. Students believed research was something you were selected for. Something that started with a professor saying yes. So they waited to be picked, and nobody picked them, and they concluded they were not the type.
The single most useful thing we did all year was tell students they could start on a Tuesday afternoon with a library database and a question they actually had. Roughly half of them had never been told that.
Recording every session mattered more than the sessions
We record every live session and post it. We did that for time zones. It turned out to matter for a completely different reason.
Students have jobs. They have siblings to pick up, shifts that move week to week, and parents who need them at home. A program that meets live and only live quietly selects for students whose time is already their own, which is the same filter we started out trying to remove.
Once sessions were recorded, completion went up among exactly the students we most wanted to reach. The flexibility was not a convenience feature. It was the access feature.
Three things we tried that did not work
We would rather write these down than pretend the year was a straight line.
- Long reading lists. We sent students five papers and they read none. We sent one paper and one specific question to answer about it, and they read it. Volume signalled seriousness to us and signalled hopelessness to them.
- Open office hours with no agenda. An empty invitation to ask anything gets no questions, because a student who does not know the field does not know what is reasonable to ask. Structured prompts fixed it immediately.
- Pairing every student with a mentor from day one. Students who had not yet written down a question had nothing to bring to a meeting, and the meeting made them feel behind. Mentors work best starting in week two, after there is something on paper to react to.
The students who finished had one thing in common
It was not prior experience, and it was not grades. It was that they picked a question small enough to be answered and stubborn enough to still be annoying them a month later.
Students who chose an impressive-sounding topic almost always stalled around week three, when the reading stopped being novel and started being work. Students who chose something they were genuinely irritated by kept going, because the irritation did not go away.
If you are deciding what to work on, that is the test. Not whether it sounds good. Whether you will still want the answer in a month.
What is next
Two things. Randy Schekman, who won the 2013 Nobel Prize in Physiology or Medicine, and Jacob Tsimerman, who won the 2026 Fields Medal, have both agreed to sit down with us for our free interview series. We are asking them the questions students actually ask, not the ones journalists ask.
And The Symposium lands in Winter 2027 in downtown Toronto, with a high school division and a university division. If you start a project now, you will have something real to present.
Everything above stays free. That was the point in October and it is still the point now.