Did you know that the question a student answers after making a mistake may be more useful than another page of the same worksheet? A learner who used the wrong relationship needs a different next question from someone who chose the right method but copied one number incorrectly. “Practise more” is not yet a teaching plan.
Choosing the next practice question after an error means selecting one small task that tests and repairs the specific decision that failed, then increasing independence and variation only when the learner can show why the method works. Start with the first unreliable step, not the last incorrect answer. The best next task is diagnostic as well as developmental.
Parents can use this approach during ordinary homework, and teachers can use it when planning a short follow-up. The eduKate Education library sets out the wider learning logic; the Clementi Secondary 1 Mathematics teaching reference illustrates why precise diagnosis should precede sequenced practice.
Find the Wrong Decision Before Picking the New Question
A marked cross tells us that the submitted answer did not satisfy the task. It does not tell us whether the learner misunderstood the relationship, forgot a routine, misread the wording, copied a value, or ran out of time. Read the working and ask a neutral question: “At what point did you decide to use this operation?” That is usually more informative than “Why didn’t you check?”
- Wrong relationship: the learner multiplies quantities that should first be compared or converted. Start with a simpler problem exposing the relationship.
- Right method, missing step: use a near-neighbour problem that demands the same step independently.
- Reading or copying slip: make the next item easy enough to isolate accurate reading and checking before raising complexity.
- Method chosen from a familiar cue: contrast two questions that look similar but require different decisions.
- Correct but slow: after verifying understanding, use brief, appropriate fluency practice rather than explaining the whole concept again.
A single error can have several causes. Treat the categories as working explanations, not labels or proof that the student is lazy, careless or naturally weak at the subject.
Worked Example: Three Pens Are Not One Pen
Consider a fictional learner answering: Three identical pens cost $4.50. What is the price of seven pens at the same per-pen price? The learner writes $4.50 × 7 = $31.50. The multiplication was carried out correctly, but $4.50 describes the cost of three pens, not one.
The missing decision is to find the unit price first: $4.50 ÷ 3 = $1.50 per pen. Then seven pens cost 7 × $1.50 = $10.50. A useful sanity check is that seven pens cost a little more than twice three pens, not seven times as much.
Now choose the next tasks deliberately. Do not simply replace 7 with 8 and ask the student to imitate the same lines. Begin with two pens cost $3; what do five cost? The unit price is still $1.50 and the result is $7.50. The task is simpler, but the learner must decide independently that the cost of the group must be divided by its size.
Next, vary the surface while preserving the relationship: four pens cost $6; what do seven cost? Again $6 ÷ 4 = $1.50, followed by $1.50 × 7 = $10.50. If the student needs no procedural cue and explains both steps, the central decision is becoming more secure.
The Contrast Question: Same Pens, Different Structure
A revealing next question is: Three pens cost $4.50, and each order has a one-off $2 delivery charge. What does an order of seven pens cost? The calculation is (7 × $1.50) + $2 = $12.50. Multiplying the full three-pen bill plus delivery by 7/3 would incorrectly scale a fixed charge that does not increase with the number of pens.
That contrast tests method selection rather than merely number processing. The student should be able to say, “The per-pen price scales with quantity; the delivery charge is added once.” Once this relationship is clear, a different setting such as equal-rate measurement can provide a later transfer check. For a real shop, discount tiers, packaging and delivery policies would have to be checked rather than assumed.
Every question and price in this example is invented for illustration. They are not actual student scripts or a commercial price claim.
A Four-Rung Practice Ladder
- Rung 1 — isolate: choose one simpler item that makes the missing idea visible. Explain it explicitly if the learner cannot yet begin.
- Rung 2 — repeat the decision, not the answer: change numbers or presentation while keeping the target method. Give the learner the first move.
- Rung 3 — discriminate: place a similar-looking but structurally different task nearby, so the learner must choose rather than pattern-match.
- Rung 4 — verify independently: use a new item later, without the model or parent prompt. Ask the learner to check the result using units, estimation, substitution or the subject’s appropriate evidence.
Move up only when the current rung produces useful evidence. If the learner cannot explain the relationship at Rung 2, do not increase difficulty merely because the homework schedule says so. Return to a diagram, concrete representation or brief worked example, then try a fresh task.
How Many Questions Are Enough?
There is no universally correct number. Two carefully chosen questions may expose a misconception better than twenty routine ones; an unfamiliar multi-step skill may require many well-spaced opportunities before it becomes fluent. The US Institute of Education Sciences practice guide describes useful learning strategies including mixing worked examples with independent problem solving and revisiting material after a delay.
A practical stopping point for one home session is when the learner can complete a suitable new task, explain why the chosen method fits, and check the answer without the previous cue—or when fatigue makes further evidence unreliable. The later follow-up is what tests retention. Extra workload is not itself a reliable measure of progress.
Three Versions of the Next Question for Different Learners
When the prerequisite is missing: if division by three itself is unstable, pause the unit-price problem. Work briefly on sharing $4.50 equally across three pens using dollars and cents, then return to the original relationship.
When the method is secure but presentation is fragile: keep the reasoning level and require the answer to show “$ per pen,” the multiplication and final dollar amount. A task designed to check units does not need to introduce a harder ratio.
When the student gets everything correct with a visible example: hide the example, let them choose the first operation, and ask for a short explanation. Supported correctness and independent method selection are different achievements; our guide to separating understanding, accuracy and speed explains why.
What to Record Across Two or Three Sessions
- Note the original error in the learner’s own working rather than rewriting it as a generic weakness.
- Record the precise teaching or hint supplied and whether the new question required the same hint.
- Keep one fresh, no-hint attempt a little later, changing the surface while preserving the underlying skill.
- Look for self-correction: the learner notices “I treated a group price as a unit price” before an adult intervenes.
- If errors return, revisit the first unreliable decision, not automatically the whole chapter.
This log need not become another homework assignment. A short sentence or marked example is enough. Teachers may already have their own assessment record; parents can communicate the observation without demanding a separate reporting system.
Mistakes That Make Practice Less Useful
Avoid selecting the next question only by page order, turning every slip into a complete re-lesson, or moving straight to exam-level mixtures when a prerequisite is shaky. Avoid keeping an answer key in view during a task meant to check retrieval. Do not use the same solved question as evidence of independent transfer, and do not interpret a learner’s hesitation as failure before they have had thinking time.
Conversely, it is possible to over-isolate a skill: endless almost-identical items can make a method appear stronger than it is. Once the core relationship is stable, questions should require the student to choose between plausible methods.
Frequently Asked Questions
Should a student redo the exact wrong question?
Yes, often once—after understanding the correction. But follow with an unseen question that tests the same decision. Copying the corrected solution from memory does not establish general understanding.
If there are five mistakes, should we repair all five tonight?
Not necessarily. Identify which error is foundational or repeatedly affects later steps. Choose a manageable priority and let the teacher know if other gaps remain. A sensible short session is better than an exhausted marathon.
Is a harder question always the best sign of progress?
No. Progress may first mean making the correct decision independently in a familiar structure. Only then is greater complexity a useful test rather than an extra source of confusion.
The Core Aim: Give the Next Question a Job
A practice question is more than something to finish. It can reveal what broke, teach one missing relationship, test method selection, or confirm that learning survived the removal of help. When adults name that purpose before choosing the next item, practice becomes more intelligent—and students become better at choosing useful practice for themselves.