Did you know a student can write a full paragraph of correct information and still fail to answer the question? Ask which school-garden plan uses less water, and a student may explain why saving water matters, how roots absorb moisture and why plants need care. All good topics. None tells you which plan to choose.
An answer misses the question when it discusses the surrounding subject but does not perform the task the question requested. The simplest repair is to underline the command, name the decision or explanation required, select evidence that settles it and end with a sentence that directly responds. This is not an invitation to write less thoughtfully; it is a way to make every sentence earn its place.
The diagnosis: what job was the question asking you to do?
Before correcting grammar or adding more facts, ask the learner to say the task aloud in plain words. Is the question asking for a comparison, a cause, a calculation, a choice, an inference or a description? Students sometimes recognise the topic—plants, fractions, historical change—yet overlook the intellectual operation.
That distinction matters because knowing facts and using them for a purpose are related but different skills. A learner who has memorised useful material may begin writing immediately on seeing a familiar keyword. If the command is “compare”, a paragraph listing everything they know about one item can feel fluent while leaving the actual comparison undone.
Worked example: two school-garden watering plans
The numbers below form an invented teaching exercise, not a report of a real school’s garden. The prompt reads: “Plan A needs an estimated 12 litres of water each day. Plan B needs an estimated 8 litres each day. Which plan uses less water, and what evidence supports your choice?”
A polished but incomplete student answer might be: “Water conservation is important for the environment. Plants need water to grow, and careful planning can reduce waste.” There is no obvious false statement. Yet the student has not compared 12 litres with 8 litres, has not chosen a plan and has not shown the requested evidence.
A direct, evidence-linked answer is: “Plan B uses less water according to the estimates given: 8 litres per day instead of 12 litres for Plan A. That is 4 litres less each day (12 − 8 = 4). If both estimates describe comparable conditions, Plan B is the lower-water option.” The response names the decision, uses the relevant values, calculates the difference and notes the comparison depends on the supplied assumptions.
If the teacher asks for the relative reduction, calculate 4 ÷ 12 × 100 ≈ 33.3% relative to Plan A. Do not add that percentage merely to sound sophisticated when the task only requests the plan and evidence. Also do not conclude that Plan B is necessarily the best gardening plan overall: water use alone does not tell us about plant survival, cost or maintenance.
A four-check audit for any written school answer
- Command check: Circle the task word: explain, compare, justify, calculate, infer or evaluate. Write the job in a short margin note such as “choose and support”.
- Evidence check: Underline the exact detail from the question, text, chart or observation needed to answer. For the garden plans, that is 12 litres and 8 litres per day.
- Connection check: Put an arrow between the evidence and the conclusion. If the reader must invent a missing reasoning step, add one sentence that explains the connection.
- Response check: Read only the final answer. Could a stranger identify which choice was made and why? If the reply still sounds like a general essay on the topic, revise its opening or conclusion.
This audit can fit on a small card. Once students know the routine, remove the card and ask them to perform the checks without prompts. The goal is an internal habit: before submitting work, they pause to ask, “Have I done what was asked?”
How question drift differs from an incorrect fact
An incorrect fact is information that does not hold. A question-drift error can contain accurate facts that are irrelevant to the requested task. Both require correction, but not the same correction. Giving a drifting student another page of topic notes may make their answer longer without making it more responsive. They need practice using known facts to complete a specific task.
Likewise, this is not identical to forgetting one subpart of a multi-part question. A student may tick every numbered item while each paragraph still avoids the main demand. Nor is it exactly the same as choosing a true but irrelevant multiple-choice option. In extended writing, the student must build the entire link between command, evidence and judgement.
Three ways otherwise capable learners drift
- Keyword-triggered recall: spotting “water” starts a rehearsed paragraph about the water cycle. Repair it by naming the required action before recalling supporting facts.
- Overloading with background: the student believes more facts must mean more marks or a better answer. Repair it by asking what each sentence contributes to the requested conclusion.
- Replacing a difficult task with an easier one: comparison feels uncertain, so the student switches to familiar description. Repair it by practising one evidence-based comparison before attempting a longer response.
A learner may also misunderstand a command word or the wording of the task. In that case, teach the meaning clearly, model one short answer and test understanding with a differently worded prompt. Some open-ended tasks permit several defensible responses, but they still require attention to the question and evidence.
Progressive practice that does not become a template
Repair stage: present a two-sentence question and three candidate answers—one directly responsive, one factually wrong and one true but off task. Ask the learner to identify the different failure modes and explain the distinction.
Stabilisation stage: give a short source with two relevant numbers, a question demanding comparison and space for a three-sentence answer. The student writes a conclusion, evidence and reasoning in whichever natural order works. Mark the function of each sentence rather than forcing every topic into identical prose.
Transfer stage: move to an unfamiliar subject, such as comparing two historical accounts or explaining why a science prediction did not match an observation. Do not tell the learner which facts are relevant. Ask them to select evidence, defend the choice and identify one limitation of the answer.
In each stage, first give feedback about whether the response has met the task. Style and sentence fluency matter, but they should not hide the more fundamental question: did the work actually answer what was asked?
A parent’s five-minute conversation after homework
When a child shows a beautifully written answer, resist the impulse to edit everything at once. Point to the question and ask, “What decision or explanation does this sentence have to deliver?” If they point to the correct part of their answer, ask for the evidence linking it to the task. If they cannot, invite a one-sentence repair, not a complete rewrite.
Then try a fresh, low-stakes question. The child’s ability to apply the audit to a new prompt is better evidence of progress than the neatness of the corrected page. Keep the original and revised responses side by side so both student and teacher can see what changed: relevance, evidence or reasoning.
What teachers can observe and measure
A quick diagnostic does not need a large marking scheme. Look for four observable features: task identified correctly, relevant evidence chosen, reasoning stated and conclusion responsive. Record which one is missing. A student who repeatedly selects suitable evidence but never states a conclusion needs a different teaching response from a student who confidently chooses the wrong evidence.
Repeat the check later with a prompt on a different topic. The aim is not to guarantee higher marks; it is to make the quality of reasoning visible and improve the match between thinking and task. Where an assignment has a formal rubric, use that rubric as the authoritative set of expectations.
Frequently asked questions
Can an answer be short and still be strong? Yes, when it contains what the task requires. Some questions need several linked points, a calculation or a nuanced explanation; length should follow the job, not replace it.
Is background information always unnecessary? No. Context may be essential to make a complex argument intelligible. The test is whether it supports the response rather than crowding it out.
What if the question is ambiguous? State a reasonable interpretation, seek clarification where possible and explain which evidence would resolve the uncertainty. Do not manufacture certainty.
Should students always repeat the wording of the question? Not mechanically. They should show they understood the task, then answer naturally and accurately.
What if the topic facts are wrong as well? Correct the content and the task alignment separately. One improvement does not automatically repair the other.
The lasting aim
The useful habit is a simple one: do not confuse talking about a subject with answering a question about it. When students can identify the requested job, choose evidence and connect it to a defensible answer, the knowledge they possess becomes more useful.
For related reading, see this series’ true-but-irrelevant answer check, the eduKate Education Hub, and the Clementi learning benchmark, which models finding the precise misunderstanding before practising and checking independently.