The format that produces diagnostic data,
not just marks.
The 3-question formative check is a structured end-of-lesson data collection instrument. Three questions in a specific sequence: one that tests recall, one that tests understanding, and one that tests application. The sequence matters. The three question types measure different things and reveal different gaps — which is why you need all three, not just the application question from A1.
The recall question is the baseline. If a student can't answer a recall question correctly, the lesson didn't land at all — they need a fundamentally different approach to the content, not a targeted fix. The understanding question checks whether students can do something with the recalled content — explain it, classify it, identify when it applies. The application question is the highest bar: can they use the concept in a new context they haven't seen before?
The diagnostic power of the 3-question check comes from the pattern across all three questions. A student who gets recall right, understanding right, but application wrong has a specific gap — they understand the concept in the contexts you taught but can't generalise it. That gap requires a different intervention than a student who gets all three wrong.
Recall. Understanding.
Application.
The recall question tests whether students can retrieve the core information from the lesson. It is deliberately low-challenge — a student who was present and paying attention should be able to answer it correctly. Its purpose is not to challenge students but to segment the class: students who can't answer Q1 need a fundamentally different intervention from students who can.
The understanding question asks students to do something with the recalled content — explain a relationship, identify why something works the way it does, classify examples. Wrong answers at Q2 reveal the specific misconceptions that need addressing tomorrow — students have the facts (Q1 was correct) but are applying them through the wrong explanatory framework.
The application question uses the novel-context format from A1. A new scenario — one not used in the lesson — that requires students to transfer the concept. Students who understood the lesson can answer this even though they've never seen this specific example. Q3 functions as both an exit ticket and a long-term progress indicator — you can reuse different versions to track the class's generalisation ability across a unit.
What the Q1–Q2–Q3 pattern
tells you to do tomorrow.
The 3-question check generates six possible response patterns. Each pattern corresponds to a specific student status and a specific instructional response. This is what makes the format diagnostic rather than merely summative — it doesn't just tell you whether students learned, it tells you specifically what didn't land and what to do about it.
The most important column is the right one. Each pattern maps to a different 5-minute opening the next lesson. The scan doesn't produce a grade — it produces a decision.
How to write a 3-question check
in under 2 minutes.
Writing three well-calibrated questions from scratch for every lesson is time-consuming. The format itself is reliable — what varies is the content. This is exactly the task AI handles well: given a learning objective and a description of what the lesson covered, an AI tool can generate a correctly structured 3-question check in seconds, with the correct answer for each question and common wrong-answer patterns to watch for.
“Generate a 3-question formative check for my lesson on [topic]. Learning objective: students should be able to [outcome]. Q1 should test recall of a key fact or term. Q2 should test understanding of the mechanism or relationship (not just recall). Q3 should require application to a novel scenario not covered in the lesson. For each question, include: the question text, the correct answer, and the most likely wrong answer and what misconception it reveals.”
This prompt produces a complete instrument with built-in diagnostic key in under 30 seconds. If you're using SprintUp Education's AI tools, the exit quiz generator handles this automatically — you provide the learning objective and the tool produces the full instrument formatted for direct use, with the misconception guide embedded alongside each question. Free on every school account.
Three subjects, three complete
3-question checks.
The following are complete, ready-to-use 3-question checks for specific lesson objectives. Each includes the questions, expected correct answers, and common wrong-answer patterns. They are not templates to copy — they are models to calibrate your own generation against.
Q1 (Recall) — What is a partially permeable membrane? Correct: A membrane that allows water molecules to pass through but blocks larger solute molecules. Wrong answer to watch for: “a membrane with holes” — reveals students are thinking physically rather than using selective permeability as the mechanism.
Q2 (Understanding) — Why does water move from a dilute solution to a concentrated one? Correct: The concentration gradient — water moves from high water potential (dilute) to low water potential (concentrated) to equalise. Wrong answer: “the concentrated side pulls the water” — reveals students have attributed active agency to the solution rather than understanding diffusion as a passive process.
Q3 (Application) — A wilted celery stick is placed in salty water. A fresh celery stick is placed in pure water. Predict what happens to each, explain the mechanism, and identify which type of solution each represents.
Q1 (Recall) — Solve: 2x + 5 = 13. Correct: x = 4. Wrong answer: x = 9 — reveals student divided by 2 before subtracting 5 (wrong order of operations).
Q2 (Understanding) — A student writes: “To solve 3x − 6 = 9, I subtract 3 from both sides first.” What mistake are they making? Correct: They're treating 3 as if it's separate from x. Correct first step: add 6 to both sides, then divide by 3.
Q3 (Application) — A box of chocolates costs €c. Buying 4 boxes plus a €3 gift bag costs the same as buying 7 boxes. Write the equation and solve for c. Correct: 4c + 3 = 7c → 3 = 3c → c = 1. Most common error: writing 4c + 3 = 7 (treating 7c as 7 rather than 7 times c).
Q1 (Recall) — Define tricolon and give an example from today's text. Correct: A series of three parallel items for emphasis. Wrong answer: vague definition without identifying parallelism as the structural feature.
Q2 (Understanding) — Why does a tricolon create a stronger persuasive effect than listing two or four items? Correct: Three creates a sense of completeness and rhythmic momentum. Wrong answer: “it sounds better” — students recognise the effect but can't explain the mechanism.
Q3 (Application) — Write a rhetorical question and a tricolon to support the argument that schools should start later. For each, write one sentence explaining the specific effect you intend.
You have the data.
Now what?
This article has given you the format that produces the most useful formative data for the time invested. But data only improves teaching when it's acted on. The final article in this cluster — A3: How to act on formative data the same day you collect it — covers the decision framework that converts your scan results into specific, targeted tomorrow-adjustments in under 5 minutes.
A3 is also where the AI-to-lesson loop closes: how to take the pattern you identified in tonight's scan and use AI to regenerate the specific part of tomorrow's lesson that needs to change — without rebuilding the whole lesson plan.