Use quiz results to group learners: a flexible decision tree for one lesson

Using quiz results to group learners can help a teacher address a shared misconception in the next lesson. It becomes risky when a short score turns into a fixed label such as “weak group” or “high ability”. A five-question quiz samples a narrow part of learning. It can suggest who needs a model, who needs another example and who is ready to apply the same idea in a new context. Those are temporary task choices, not identities.

The Education Endowment Foundation’s review of setting and streaming concerns persistent grouping across classes, not a ten-minute task group. It nevertheless warns about misallocation and recommends flexibility. An EEF explanation of flexible grouping describes grouping around current needs. The inference here is limited: use a quiz to plan a short intervention, then move learners as new evidence appears.

Ask a question that distinguishes next steps

Suppose a class is learning to compare fractions. A quiz asks whether 3/5 is greater than 3/8 and requires a one-sentence reason. Learners might choose the right fraction because they understand that fifths are larger than eighths when numerators match. Others might choose it but say “5 is larger than 8”, revealing a contradiction. Some choose 3/8 because they compare denominators as whole numbers. A correct letter alone is not enough to decide what each learner needs.

Add a second item with a diagram or a different numerator. If the two results disagree, ask a quick oral follow-up rather than forcing a learner into a group. The diagnostic quiz guide explains how to connect an item to a prerequisite. This grouping example starts only after the answer key and wording have been checked.

Record the objective and the response pattern, not an overall rank. “Compared denominators as whole numbers on item 2” is actionable. “Scored 40%” is not. Keep the notes minimal and protected under your school practice; there is no need for a public list of names or permanent labels.

Use a three-way decision tree

After reviewing responses, choose one of three short activities. The groups can sit in different places, but they do not need labels visible to the whole class. Tell everyone that the tasks reflect today’s question and that movement is expected.

Evidence today	Task for ten minutes	Check afterwards
Denominator treated as a whole number	Build two equal-sized fraction strips and explain piece size	New comparison with the same numerator
Correct choice, weak explanation	Pair two written reasons and improve one	One-sentence justification
Correct choice and sound reason	Solve a comparison with unlike numerators	Explain method and its limits

This is a decision tree, not an automatic assignment rule. If a learner solved the first item incorrectly because the drawing was unclear, they may not need the strip activity. If someone in the extension group struggles with the new problem, move them into a short discussion. The point is to match a task to a current need and collect new evidence quickly.

A group can also be mixed when explanation is the goal. A student who articulates piece size clearly may help another by comparing reasoning, provided the conversation is structured and the same students are not always made unpaid tutors. The teacher remains responsible for teaching each group. Do not reserve the richest explanation for one group only.

Check whether the grouping helped

After the ten-minute activity, ask everyone a parallel question: “Which is larger, 4/7 or 4/9? Explain.” The numbers changed but the core comparison is similar. Read the reasons, not just the answer. If a learner now describes the size of sevenths and ninths accurately, they can move on. If the answer is still unstable, offer another representation or a one-to-one conversation. An exit ticket can provide a later check. The What Works Clearinghouse study guide supports reviewing important material after a delay, although it does not endorse this exact grouping scheme.

Compare the pre- and post-question cautiously. The second item may be easier, and students have just heard the answer discussed. A delayed question in a different context helps show whether the idea holds. Use the results to revise tomorrow’s instruction rather than to prove a treatment effect from one lesson.

The EEF feedback guidance asks teachers to address identified gaps with clear next steps. For the denominator misconception, “think harder” is not feedback. “Show why four sevenths cover more of an equal whole than four ninths” gives the learner an action they can take.

Protect learners from fixed labels

Avoid public group names like “bottom set”. Even neutral names can become codes if the same students always receive the same work. Explain the task rather than the category: “This table is using diagrams; this one is improving explanations.” Make movement ordinary and frequent. A learner can need a visual model in fractions and be confident in another topic. A five-question result does not measure general mathematical ability.

Watch whether the group work actually gives each learner access to high-quality teaching. If one group spends the whole time on repetitive low-level items while another discusses ideas, the plan may deepen a gap. The EEF’s evidence on persistent setting is not a direct estimate for short groups, but its caution about assignment and access to teaching is worth considering. Have a clear route from each task to the same learning goal.

Check accessibility and participation too. If the quiz required fast reading on a phone, a slow response may reflect the format. Offer another way to explain a reason before assigning a task. If a student was absent, use a brief question at the start of the lesson rather than treating missing data as low attainment.

Build the cycle in Skolina

Create two or three focused questions in Skolina, share the quiz by link or QR code and review responses by question. Objective items can be marked automatically; read the short reasons yourself. Use the distractor guide to make each wrong option informative. Check current plans if you need several versions of the same check.

To try one lesson, create a Skolina teacher account, choose one misconception and prepare three ten-minute tasks before sending the quiz. After the lesson, ask a parallel question and change the groups where the new reasons warrant it. That movement is a sign that the quiz is informing teaching.