Quizzes in the classroom: why they work, how to write them, and how to run one for thirty students

The best-replicated finding in the psychology of learning is that trying to recall something strengthens the memory of it far more than looking at it again does. The classic experiment, run many times since Roediger and Karpicke published it in 2006, gives one group a text to study twice and another the text once followed by a test, and a week later the tested group remembers substantially more — even though the studied-twice group felt more confident at the time. Recalling is the learning; re-reading only feels like it.

A quiz is retrieval practice with the awkwardness removed. Nobody sits alone with a blank page; there is a clock, a leaderboard, and the small social pressure of a room. The score matters less than the act of pulling the answer up, and a wrong answer followed by the right one on the screen is, for memory, nearly as good as a right answer — the correction lands exactly where the gap was.

Two things follow for teaching. First, the quiz belongs after the lesson, not instead of it: it consolidates what was taught, it does not introduce it. Second, low stakes are the point. A quiz that counts toward a grade makes students avoid guessing, and a student who does not guess does not retrieve.

The five-minute opener. Five questions on last lesson's material, before anything new is taught. It is the single highest-value use of a quiz: it reactivates what was covered a day or a week ago, right before it is built on. Keep it fast, keep it easy, and do not discuss the answers beyond a sentence each — the retrieval has already done its work.

The exit ticket. Five to ten questions on today's lesson, in the last ten minutes. This one is for you as much as for the students: the answer counts on each question show which idea landed and which did not, before the class has left the room. A question that three quarters of the class got wrong is the first thing to say tomorrow.

The revision round. Fifteen to twenty questions across a unit, in the week before a test, played in teams so the discussion does part of the work. Mix difficulty deliberately and put the hard questions last, so the round tightens rather than opens hard and discourages the weaker half of the room. This is also the format where estimation questions earn their place: "roughly how many years did the Thirty Years' War last" is a question about the shape of the period, not a date.

A multiple-choice question is only as good as its wrong options. Three random distractors produce a question anyone can guess by elimination; three plausible ones produce a question that tests the distinction you actually taught. The best source of plausible distractors is your own marking: the wrong answers students actually give are, by definition, the confusions worth surfacing. A physics question whose wrong options are the three most common unit errors is a lesson in itself.

Ask about the thing, not the wording. "Which of the following statements is true of photosynthesis" tests reading; "what does a plant take from the air in photosynthesis" tests photosynthesis. Keep the question stem short and put every word that matters at the front — under a clock, the last clause of a long sentence is the one that gets missed.

One question, one idea. A question that needs two facts to answer is two questions, and when it goes wrong you learn nothing about which fact was missing. And write the answer options at the same length: students learn quickly that the longest option is usually the right one, because it is the one the teacher took care over.

On omequiz a quiz is written in the editor: title, questions, four answer options each, one correct, optionally an image per question. It is saved as a draft, so a quiz can be built across several evenings and played as often as needed. The questions can be written in German, English or both.

A room holds up to ten players. For a class of thirty, that means three rooms — which is also how many teachers run it anyway, as three teams whose scores are compared at the end. The host of each room does not have to play: in quiz-night mode the host runs the round from a projected screen, reads each question, watches the counter of how many have answered, reveals when the room is ready and moves on. The students answer on their own phones, and the clock closes each question after twenty seconds — enough to answer, not enough to search.

On accounts: playing in a room needs an account, which is created with a Google sign-in. A quiz that has been published can also be played alone, without any account, by anyone who has the link — which is the simplest way to hand a revision round to a class as homework. There are no advertisements on the game screens; ads appear only on reading pages like this one.

The leaderboard is for the room; the per-question numbers are for you. After a round, the questions with the lowest correct rate are the list of what to teach again, in order. Do it the next lesson, not next month — the value of an exit ticket decays fast.

And resist making the leaderboard permanent. A class-wide ranking that persists across weeks turns a learning tool into a status contest and switches off exactly the students who need the retrieval most. Each round stands alone, the score is forgotten by tomorrow, and the only thing that carries over is what was remembered.