Ask a class "does that make sense?" and almost everyone nods. Ask the same class the same question without names attached and you routinely find a third of the room saying no.
Both answers are real. The first is what people say when their name is attached; the second is what they think. Whether the room can see who said what is the switch between them, and knowing when to flip it is most of the skill.
What a not-shown poll actually buys
It removes the social cost of an answer. In a classroom, admitting confusion carries a price — looking slow in front of peers, disappointing a teacher who just spent twenty minutes explaining, being the one person who did not get it. Students pay attention to that price and answer accordingly, which means a named "did you understand?" measures social confidence rather than understanding.
Take that visible price away and the answer changes. That is the entire mechanism, and it is why a poll is so much more informative than a show of hands. It is also why the no-names "did that make sense?" is one of the five-minute formative assessment ideas worth keeping.
What it costs
Follow-up, and all of it at once. An unattributed answer cannot be marked and cannot tell you who needs help — only how many. You learn the size of the problem and lose the ability to act on it individually.
There is a second, subtler cost: it removes the incentive to think carefully, since nothing follows from the answer. For an opinion that is fine. For anything where effort matters, it is a real problem.
The test: would you act on this answer differently if you knew who gave it? If yes, ask it as a normal question. If no, a poll is free information.
Use a poll for these
Comprehension checks
"How well did that explanation land?" with a four-point scale. You are deciding whether to move on or reteach, and that decision depends on the count, not on the names.
Pace
"Too fast / about right / too slow." Almost nobody will tell you this to your face, and almost everybody has an opinion. It is one of the highest-value questions a teacher can ask and one of the least asked.
Prior knowledge, before you teach
"Have you met this idea before?" Asked with names, students under-report to avoid seeming arrogant and over-report to avoid seeming behind. Asked as a poll you get a usable distribution, and you can pitch the lesson at it.
Opinions with no right answer
Ethics discussions, debate positions, "which of these do you find most convincing". The point is the spread of views, and attaching names narrows the spread immediately as students converge on the socially safe position.
Feedback on your teaching
"What would make this module better?" You will get more honest and more useful answers when the room cannot see who wrote what — provided students believe you, which is a claim you have to earn and should not overstate.
Do not use them for these
- Anything graded. Obvious, but worth saying: a poll is not scored, so a mark cannot come out of one.
- Anything you intend to follow up individually. "Who is struggling with question four" is useless as a poll. Ask a normal question instead and read the answers.
- Anything requiring real effort. A question with no name on it gets no-name-on-it effort.
- Safeguarding or wellbeing disclosures. Genuinely dangerous — it can surface a serious concern with no reliable way to identify and help the student who raised it. Use your school's proper channel, never a classroom poll.
- Anything confidential. Not showing names is a presentation choice, not a security guarantee. Do not invite an answer a student would be harmed by having traced back to them.
- Small groups. In a class of six, not showing names is a polite fiction. Everyone can work out who said what, and pretending otherwise damages trust more than not asking would have.
Asking them well
Say what will happen to the answers
"I'm going to show these results in a second and we'll decide together whether to move on." Students answer more honestly when they can see the point, and more cynically when a poll looks like a formality.
Act on it visibly, at least sometimes
The fastest way to kill honest responses is to ask "too fast?", get a clear answer, and carry on at the same pace. Once that happens the class correctly concludes the poll is decorative.
Avoid the agreeable middle
A five-point scale collects a pile of 3s. Four points forces a lean. If you want a decision out of the poll, do not offer an option that means "no opinion".
Keep them rare
A poll every lesson becomes routine, and routine questions get routine answers. Two or three a term, at genuine decision points, produce far better data than fifteen.
Do not show a live-updating bar chart
If results appear as people vote, later voters follow earlier ones. Collect first, reveal after. This one is easy to get wrong and it quietly invalidates the result.
The combination worth knowing
The most useful pattern is not a poll on its own. It is a normal, attributed question followed by a confidence poll: ask the content question so you know who got it right, then ask "how sure were you?" as a poll.
The cell you are looking for is confident and wrong. Those students will not ask for help, because they do not believe they need it, and ordinary re-teaching slides straight past them. You cannot find that group any other way.
Quizoma runs polls alongside scored questions in the same quiz. Polls are never scored, and with anonymous voting switched on the app shows totals rather than who picked what — treat that as a presentation choice rather than a privacy guarantee. See the question formats.