What makes students speak up isn't class size. It's whether it's safe to be wrong.

The finding

There is a common assumption that big classes silence students. The data complicates it. Teacher Interaction, whether students can interact with and reach their teacher, is the single highest-rated thing students report in the whole dataset: 9.1 out of 10 across roughly 1,360 responses. Participation Confidence and Class Participation sit nearly as high, at 8.6 each. Almost nobody rates interaction badly. Only 1% land in the bottom band, while 89% land in the top.

If class size were the lever, we would expect a wide spread of scores tracking how crowded the room is. We do not see that. Interaction stays high, and the rare low scores cluster around something else entirely.

What the low scores reveal

Read the 1% who rate interaction badly and a clear pattern emerges. What silences students is not the size of the room. It is psychological safety, whether it is safe to be wrong. Students go quiet when asking a question gets them belittled, laughed at, or met with visible irritation. Where interaction thrives, students describe the opposite: a teacher who invites questions, makes it safe to be wrong, and balances talking with asking.

Where students feel unsafe to speak 2.5

"If you say something wrong, they don’t try to help. They belittle you."

"There’s always someone who’ll make a dumb comment or laugh at you if you ask about something they think is obvious."

"I’m just not comfortable speaking in front of people."

Where interaction works 9.5

"You can ask questions without anyone saying anything annoying or calling it dumb."

"A good dialogue and questions in class, and a teacher who balances talking with asking."

"For every new topic there’s a toolbox full of parts you can handle, so you understand the material better."

Representative, de-identified comments. Translated from the original where needed.

Why the reframe matters

The seating count is not what students respond to. The climate is. A small class run by a teacher who lets students feel exposed for asking will still go quiet. A larger cohort with norms that protect the student who risks being wrong will keep talking. The institutional implication follows directly: as cohorts grow, protect the conditions for safe interaction, because that, not the number of chairs, is what students actually respond to.

What it means for institutions

1. Stop treating participation as a class-size problem. The data says interaction stays high even as the norm, so the fix is climate, not just smaller rooms.
2. Make psychological safety an explicit teaching standard. “No question is stupid”, no belittling, visible patience. It is what separates the 1% from the 89%.
3. As cohorts grow, design for safe interaction deliberately. Small-group time, anonymous question channels, and norms that protect students who risk being wrong.

How we measure it

Teacher Interaction sits in the Social domain under Class Size & Interaction, with a related read on Confidence & Self-Efficacy. Students rate the experience on a 0 to 10 scale, here aggregated across roughly 1,360 responses, and we report the mean alongside the distribution across bands (bottom 0-4, top 8-10) so a high average cannot hide a struggling minority. Open-text comments are de-identified and, where needed, translated from the original. The quotes shown are representative of the low- and high-scoring segments (segment means of about 2.5 and 9.5), not cherry-picked outliers.