Alicia reaches the answer and puts down her pen. The difficult part is over. Her equation is appropriate, the substitution is correct, and the calculator has returned a sensible number. On the answer line she writes the number exactly as it appears on the screen. The question, however, asked for the minimum number of complete containers required. A decimal is an intermediate result, not the requested decision.
Tricia encounters a quieter version of the same problem. She calculates the difference between two values correctly but writes the larger value as her conclusion. Kai Kai solves an equation accurately and keeps both algebraic roots, although one is impossible in the situation. Their work contains substantial success. Their finished responses do not yet deliver everything the question requires.
Alicia, Tricia and Kai Kai are fictional learners. Their examples are original teaching illustrations, not reports of measured student outcomes. Throughout this article, “the final mark” means the credit attached to completing an answer correctly; it is not a claim that every examination reserves exactly one mark for the last line.
The useful question is not why examiners refuse to appreciate hard work. It is why a learner can finish the reasoning they planned and still fail to finish the task they were given. The repair begins by treating the final answer as a decision of its own.
1. The 50-second route: check the result against the request
Before submitting an answer, reconnect the result to the original instruction. Name the quantity requested. Check the required form, precision, units and conditions. Confirm that the response has been entered in the correct place. Then stop unless a specific contradiction remains.
This is not a command to redo the entire solution. Repeating correct calculations will not reveal that the question asked for a percentage rather than an amount, a coordinate rather than a distance, or a conclusion rather than a description. Completion checking looks outward from the calculation to the task.
A useful spoken rehearsal is: “I found this quantity. The question wants this quantity. My final response connects them.” At first a student may say it aloud during practice. Later it can become a brief mental check. The routine should shrink as it becomes reliable.
There are also occasions when the working is not actually correct. A familiar formula may have been used in the wrong situation. A valid calculation may answer a different question. A response may merely resemble the model answer. In those cases, a final-line routine is not enough; the student needs to revisit the earlier decision.
The distinction matters because two opposite mistakes waste time. One student repeats an entire chapter to fix a response-format problem. Another keeps adding units to an answer whose model is wrong. First establish where the route stopped being valid, then repair that part.
2. Understand the marking system without inventing one
A student should not learn marking rules from a universal slogan. Some assessments credit intermediate reasoning. Others score an item as correct or incorrect. Even within one paper, different questions may use different criteria. Units may be required in one response, supplied beside another response box, or treated according to qualification-specific instructions.
The safe source is the relevant question paper, mark scheme, examiner guidance and teacher explanation. These materials tell the learner what was expected in that assessment. They do not necessarily establish rules for every other subject or examination board.
For a concrete example, AQA’s GCSE Statistics command-word guidance distinguishes exact values, required accuracy and completing a table. It also explains that answers entered in a table take priority over answers in working spaces for that command. That is a specific published convention, not a reason to assume every examination uses identical answer-priority rules.
The learning lesson is broader: answer placement and requested form can be part of the assessed task. Students should know those requirements before the examination, rather than discovering them through a disappointing result.
When a mark seems questionable, preserve the original response and ask which criterion was not met. A teacher may confirm that the marking was correct, identify an acceptable alternative, or find a marking error. Good examination preparation allows that conversation. It does not teach unquestioning acceptance of every red cross, and it does not assume every lost mark is unfair.
3. A calculation can be correct and still stop one decision early
Consider this original problem. A group needs to transport 137 passengers. Each vehicle has space for 24 passengers. How many vehicles are required if everyone must travel?
The division gives 137 ÷ 24 = 5.7083… . Alicia has performed the division correctly. Writing 5.7 vehicles does not complete the task, because the decision concerns whole vehicles. Five vehicles provide only 120 seats; six provide 144. The answer is six vehicles.
Now change the problem. A workshop has 137 metres of material. Each complete banner requires 24 metres, and offcuts cannot be joined. How many complete banners can be made? The same division appears, but the answer is five complete banners. Six would require 144 metres, which are not available.
The arithmetic is identical. The finishing decision differs. One problem asks for enough capacity; the other asks how many complete products fit within a limit. Ordinary rounding to the nearest whole number is not a substitute for reading the situation.
This pair is useful precisely because it prevents students from memorising “always round up in word problems.” Ask them to explain what would go wrong with the other integer. Their explanation should mention capacity or available material, not the final decimal digit.
A learner who can make this distinction has improved the connection between mathematics and the world. A learner who merely writes the expected integer after being told the answer has corrected one exercise. The next check should use a different context.
4. Track the noun attached to each number
Many finishing errors are changes of quantity disguised as copying mistakes. A student calculates cost, discount, tax, final price and saving in the same solution. The final line then selects the wrong number from a page full of correct arithmetic.
Take a price of 240 currency units reduced by 15%. The reduction is 36, and the new price is 204. If the question asks for the amount saved, 204 is wrong. If it asks for the sale price, 36 is wrong. Neither error requires reteaching percentage multiplication. The learner needs to preserve the identity of the quantities.
During practice, label intermediate values with short nouns: “reduction = 36”; “new price = 204.” Those labels make the last decision easier. They also help when a multi-part question asks the student to reuse one quantity rather than another.
The same principle applies to science. Mass, change in mass and percentage change in mass are different outputs. So are final temperature, temperature increase and rate of temperature change. A correct measurement difference is not automatically the requested rate.
Do not require elaborate prose beside every calculation. The purpose is to prevent quantity identity from disappearing, not to make working decorative. A concise label at a risky transition can be enough.
For a student who repeatedly loses the last mark, examine whether the wrong answer is already sitting somewhere else in the correct working. If it is, the primary repair may be selection and labelling rather than calculation.
5. Exact form and approximate form do different jobs
Suppose a question asks for the exact area of a circle with radius 3 units. The mathematical expression is 9π square units. A decimal such as 28.27 is an approximation. It may be an excellent numerical estimate, but it does not meet an explicit request for an exact expression.
Now suppose another question asks for that area to two decimal places. The response should be 28.27 square units. Leaving 9π alone may not complete the requested presentation. The mathematics and the response instruction must agree.
The word “exact” does not mean “write many decimal places.” A long calculator display can still be approximate. Conversely, a short fraction or expression can preserve the quantity exactly. Students need this conceptual distinction before they can reliably check the final form.
OCR’s explanation of mathematical accuracy discusses exact and rounded answers and warns against both excessive and insufficient specification. Its qualification context matters, but the distinction helps explain why presentation is sometimes mathematical rather than cosmetic.
Use paired finishing tasks. Give the same completed calculation with two different final instructions, one requesting exact form and one requesting a stated accuracy. Ask the learner to produce both endings and explain why they differ.
This is more targeted than another page of calculating areas. The area calculation was already secure. The missing step was choosing the form in which the result should be communicated.
6. Keep intermediate precision separate from final presentation
Premature rounding is an upstream error that often appears only at the end. A learner rounds an intermediate value, uses that rounded value in later calculations and produces a final result outside the required tolerance. The final line is where the difference becomes visible, but the first loss of information happened earlier.
For a simple illustration, suppose an intermediate quantity is 0.146 and must be multiplied by 300. Keeping it gives 43.8. Replacing it with 0.15 first gives 45. The multiplication can be executed perfectly in both routes, yet the second route answers a modified problem.
The practical habit is to retain sufficient intermediate precision and apply the requested rounding to the final result. A calculator’s stored value can help, where calculator use is permitted. A student should still understand what is being stored rather than pressing a sequence of buttons without meaning.
Do not teach that every intermediate value must always have a fixed number of decimal places. That shortcut can fail for very small or large quantities. The precision needed depends on the calculation and the assessment instructions.
The detailed rounding lesson already has a place in the wider library: How to Use Rounding and Significant Figures in Exams. Here the narrower job is diagnosis: was the final response merely formatted wrongly, or had information already been lost during the route? The two cases require different repairs.
7. Units identify what the number means
A number without its quantity can be ambiguous. Twelve metres, twelve square metres and twelve metres per second describe different things. A unit is not an ornamental suffix; it can expose whether the learner has found the intended kind of result.
Consider a distance of 900 metres travelled in 3 minutes. The rate is 300 metres per minute, or 5 metres per second. A student who divides correctly and writes “300 metres per second” has attached the wrong meaning to a correct quotient. The issue is not simply forgetting a unit. It is failing to track the units used in the division.
Another learner calculates an area of 0.24 square metres and converts it to 24 square centimetres. The linear conversion has been used where an area conversion is required. Since one metre is 100 centimetres, one square metre contains 10,000 square centimetres; the area is 2,400 square centimetres.
These examples show why “remember your units” can be too weak a correction. Ask the student to explain how the unit changes through the calculation. That explanation identifies whether the weakness is notation, conversion or understanding of the quantity.
Actual mark consequences remain scheme-dependent. Do not invent a universal one-mark penalty. The educational priority is to make the response meaningful and compliant with the question. When a unit is already printed beside the answer box, students should complete the box as instructed rather than duplicating or contradicting the supplied unit.
8. Candidate solutions are not always accepted solutions
Solving an equation can generate candidates that still need checking against the original conditions. This is particularly visible after operations such as squaring, which can remove a sign distinction.
Consider √(x + 6) = x. Squaring gives x + 6 = x², so x² − x − 6 = 0 and the algebraic candidates are x = 3 and x = −2. Substitution into the original equation confirms 3, because √9 = 3. It rejects −2, because √4 = 2, not −2.
The factorisation was correct. The missing finishing step was validation. Calling both candidates “the answers” ends the process too early. The learner must distinguish solving a transformed equation from satisfying the original equation.
Context can impose a different kind of restriction. A model might produce two times, one before the stated observation period and one inside it. A geometric calculation may produce a negative length candidate. The valid response must respect the situation described, not merely the algebraic manipulation.
Practice should include cases where all candidates are valid as well as cases where one is rejected. Otherwise students may learn a new superstition: always cross out one root. The reason for rejection must be explicit.
A short validation note is often enough in working: “−2 does not satisfy the original equation.” The final answer should then clearly identify the accepted solution. This is mathematical completion, not a request for elaborate presentation.
9. Finish the set, interval or relationship requested
Some students solve almost everything correctly but report only part of the solution. The problem is not a wrong number; it is an incomplete set.
Suppose x² = 49 over the real numbers. The solutions are 7 and −7. Writing only 7 loses the distinction between the principal square root √49 and the solutions of an equation involving x². That distinction belongs to the mathematics itself.
For an inequality, −2x > 6 leads to x < −3 after dividing by a negative quantity. A final response of “−3” names a boundary, not the values satisfying the inequality. The answer needs the relation as well as the number.
Coordinates create another completion demand. Finding x = 4 is not enough when the question asks for the intersection point and the corresponding y-value is required. A distance or gradient is not a coordinate pair either, even if all three are calculated in the same solution.
Ask students to name the expected answer object before they start: a number, pair, interval, set, expression, diagram or sentence. This helps them recognise when the work has not yet produced that object.
The habit is especially useful for multi-part problems. A student may carry the answer form from the previous subpart into the next one. Reading the command again at the finishing point prevents the old output requirement from silently replacing the current one.
10. Preserve the reference in comparisons and percentages
A correct difference can still support a wrong final claim when the reference is lost. If a value rises from 40 to 50, the increase is 10 and the percentage increase relative to 40 is 25%. The increase is not 20% merely because 10 is one fifth of the final value.
Likewise, a success rate moving from 60% to 75% rises by 15 percentage points. Its relative increase is 25%, because 15 is one quarter of the starting 60. These are different descriptions of the same change. The question determines which one is wanted.
Students often calculate several of these values correctly in scratch work and then choose a familiar phrase for the conclusion. That phrase can change the meaning. The final sentence should therefore preserve both the quantity and the comparison base.
An effective finishing prompt is: “Compared with what?” Use it for larger, smaller, faster, more efficient and other comparative claims. A statement that one material loses less heat is incomplete if the comparison object and conditions are not clear in the answer context.
Do not make students write redundant reference phrases when the response format already supplies them. The point is meaning, not verbosity. A short table entry under a clearly labelled heading can be sufficient. A standalone conclusion may need more words because the reference is not carried by the table.
11. Science answers must connect evidence to the requested conclusion
Correct scientific observations do not automatically constitute a complete explanation. A learner may report that one object cooled faster, another stayed warmer and a third had a different surface. Those observations can be accurate while the causal relationship requested by the question remains unstated.
Consider an original classroom scenario: two otherwise comparable containers begin with water at the same temperature. One has insulation. After a specified interval, the insulated container’s water is warmer. If asked which container lost less thermal energy under the stated conditions, the answer should identify the insulated container and connect the temperature evidence to the conclusion, subject to the assumptions supplied in the task.
Writing only “it was warmer” may leave the referent unclear. Writing “insulation creates heat” adds an incorrect explanation. Writing an entire paragraph about every heat-transfer mechanism may obscure the simple comparison the question actually requests.
The completion check asks: have I named the object, used the relevant evidence and supplied the relationship required by the command? If the command is only to state an observation, do not demand a causal essay. If it asks for an explanation, an observation alone may not be enough.
For students who repeatedly lose these marks, separate evidence selection from explanation production in practice. First choose the relevant data. Then write the missing relationship. Finally return to a fresh complete question without those stages being labelled. The goal is independent completion, not permanent dependence on a teacher’s sentence frame.
12. Language answers can stop before their communicative job is done
In English and humanities, the equivalent of the missing last line is often an unfinished relationship. A student selects a relevant quotation but never explains what it shows. Another makes a reasonable claim but does not connect it to the question’s focus. The writing is related to the task without fully answering it.
Take this original sentence: “When the invitation arrived, Nora put it beneath the unopened bills and changed the subject.” Asked what this suggests about her response to the invitation, a student might quote the action accurately. The next step is a cautious inference, such as reluctance to deal with it, supported by concealment and changing the subject. A claim that she hates every social event would extend beyond the evidence.
Completion therefore involves both adding and limiting. Add the relationship that is missing. Stop before inventing motives the passage does not support. More writing is not automatically more complete.
In argumentative work, the final sentence of a paragraph should often explain why the evidence matters to the claim, not simply repeat the evidence in different words. But no universal paragraph formula governs every task. Follow the purpose and criteria.
A practical exercise is to compare three responses: relevant evidence only, evidence plus a supported conclusion, and evidence plus an exaggerated conclusion. Ask the learner which one finishes the job and why. The discussion trains sufficiency without turning writing into a checklist of fashionable phrases.
13. The transfer from working space to answer space is a real operation
Some losses happen after the intellectual work is complete. A learner has the correct result in the margin but copies a different value into the answer box. Another types a decimal point incorrectly. A third calculates the correct response for one row and enters it in the next row.
Treat that transfer as an operation worth checking. The source is the result the student intends to submit. The destination is the required answer location. The task is to preserve the value, sign, unit convention and question identity across that move.
In practice, ask students to point to the source and destination on a few selected items. They need not draw arrows all over an actual examination paper. The exercise simply makes the hidden operation visible.
For digital assessments, rehearse the available practice interface when provided. Mathematical entry can require particular notation. A box may already contain a unit. A table may require one value per cell. Submission and navigation behaviours differ, so do not assume the rules of one platform apply to another.
Preserve permitted accessibility supports throughout this work. A learner using an approved keyboard, enlarged display or other access arrangement is not showing weaker independence merely because the interface differs. The relevant question is whether the response is produced and submitted accurately under the assessment conditions that legitimately apply to that learner.
14. Make the final answer unambiguous without erasing the evidence
A page can contain several possible endings. One is an intermediate quantity, another a rejected candidate, and a third the intended answer. A marker should not have to guess which one the student means.
During practice, teach clear correction and labelling in the format the examination permits. Cross out a replaced answer as instructed, keep necessary working readable, and identify the final response. Do not create a page on which incompatible answers appear equally endorsed.
This does not mean every additional line automatically cancels previously earned marks. Marking conventions differ, and some schemes explicitly address subsequent working. The safer educational habit is clarity rather than speculation about which contradiction a marker might ignore.
A learner who notices a genuine error should correct it. A learner who merely feels uneasy should not generate a second answer for reassurance. The companion guide on changing multiple-choice answers develops the evidence threshold for reconsideration. Here the concern is leaving a single intelligible submission after that decision.
Clear finalisation also helps the student. When returning to a question later, they can see what remains unresolved. An answer labelled “candidate; check original condition” is easier to resume than an unexplained collection of numbers. Good working is a record of decisions, not merely a record of arithmetic.
15. Use a finishing check, not a second full solution
The finishing check should be short enough to survive a real paper. A useful sequence is: requested quantity, required form, applicable condition, answer location. Units and precision fit inside required form, while physical restrictions or original-equation checks fit inside applicable condition.
Not every question needs equal attention to every category. A direct vocabulary response may need a meaning and grammar check, not a unit check. A numerical model may need all four. The learner should select the checks that fit the actual task and their known error pattern.
Tricia’s previous routine was to repeat the calculation. It sometimes caught arithmetic errors, but it missed the fact that she had answered with the larger amount instead of the difference. Her new routine begins by rereading the noun phrase after “find.” The changed check targets the actual failure.
Kai Kai’s routine focuses on candidates and conditions. Once he has generated solutions, he returns to the original equation or model before accepting them. Alicia’s routine focuses on the move from a decimal quotient to a whole-number decision.
These are not three personalities requiring three permanent labels. They are three temporary error patterns. If fresh work shows the pattern has changed, the checking routine should change too. A useful check earns its place by preventing or detecting a relevant error; it should not become an endless ritual maintained out of habit.
16. A six-item finishing workshop
Use this original workshop after the relevant concepts have already been taught. The learner receives working that is mostly complete and must finish the answer. This isolates completion without requiring another long calculation on every item.
First, 83 visitors need minibuses holding 12 people each. The calculation gives 6.9166… . Finish with seven minibuses, and explain that six hold only 72 people. Second, 83 metres of ribbon are available for 12-metre pieces that cannot be joined. Finish with six complete pieces, because seven require 84 metres.
Third, a calculation gives 0.04678 and the instruction asks for three significant figures. Finish with 0.0468. Ask the learner to distinguish this from three decimal places, which would give 0.047. Fourth, an exact result is 7/12 and the instruction explicitly asks for a fraction in simplest form. Keep 7/12 rather than replacing it with a rounded decimal.
Fifth, the equation (x − 1)² = 16 has been reduced to x − 1 = ±4. Finish with x = 5 or x = −3, unless a stated context imposes an additional restriction. Sixth, a journey covers 1,200 metres in 4 minutes and asks for metres per second. Finish with 5 metres per second, not 300 metres per second.
These are mathematical completion answers, not an invented official mark scheme. The teacher can assess whether the learner selected the correct quantity and form without claiming that every examination would allocate the same marks to each response.
17. Move from isolated finishing to fresh complete questions
A finishing workshop can become too helpful if the learner is always told that only the ending needs attention. In a real paper, the error may occur anywhere. The next stage must therefore remove that cue.
Give fresh complete questions containing a mix of demands. Some require a whole-number decision, some exact form, some no special finishing transformation. The learner solves normally and decides which checks are relevant. Include ordinary questions where the first numerical result is already the final answer, so that the student does not learn to transform every result unnecessarily.
Then return after a delay with another small set. Change numbers and context. Do not prove mastery by asking the learner to repeat the six workshop answers from memory. Success should reflect the finishing decision, not recollection of the examples.
Eventually inspect a full practice paper. Did the relevant omissions fall without completion time becoming worse? Did the student preserve correct earlier reasoning? Did they stop adding unnecessary corrections? These are more useful questions than whether the new checklist was used on every item.
A checking routine that catches two errors but consumes the time needed for several unanswered questions may need redesign. The goal is better whole-paper performance. Finishing must protect the answer without taking over the examination.
18. Diagnose whether the missing mark is a small repair or a deeper gap
A learner may lose the same visible mark for different reasons. One student omits the negative root because they forgot to record it. Another believes square equations have only positive solutions. The first may need a completion habit. The second needs conceptual teaching.
Use a contrast question before prescribing the repair. Ask why both 7 and −7 satisfy x² = 49. If the learner explains accurately without help, inspect the recording habit. If not, return to the relationship between squaring and solving. The location of the lost mark does not by itself identify the size of the knowledge gap.
The same distinction applies to units. A student who knows that square metres convert differently from metres may simply have copied a linear unit. A student who cannot explain the conversion needs more than a reminder. Likewise, a sentence that omits a causal link may reflect hurried writing or missing understanding of the mechanism.
This is where How Learning Diagnosis Works becomes useful. Preserve the distinction between finding the visible defect and explaining its cause. The first tells you what to inspect; the second tells you what to teach.
When several connected prerequisites are weak, a broader lesson is appropriate. Surgical repair does not mean refusing to reteach foundations. It means making the scope of the repair proportional to the evidence rather than automatically restarting everything or automatically treating everything as a slip.
19. Keep the record small and the original score honest
A useful record can contain four fields: the original response, the requirement missed, the finishing action needed, and the result on a fresh check. There is no need to preserve every question forever.
For Alicia, an entry might read: “Decimal number of vehicles; needed whole vehicles sufficient for everyone; test capacity before rounding; fresh transport question completed correctly.” For Kai Kai: “Kept transformed-equation candidate; needed validation in original equation; substitute candidates; fresh radical equation still needs practice.”
These entries are specific enough to guide revision. “Careless” would not tell either student what to do differently. A long copied solution may also be less useful than a short statement of the decision that failed.
Keep corrections separate from the original mark. A student who adds the unit after opening the scheme has improved the answer, not changed what was submitted. Use the guide on self-marking past papers without inflating the score when those stages are becoming mixed.
As a finishing habit becomes stable across fresh work, reduce special monitoring. Retain occasional checks within normal practice. The record should help a learner become less dependent on reminders, not create a permanent administrative system around an error that no longer occurs.
20. Read the answer from the marker’s side
A useful rehearsal is to hide the working temporarily and read only the question and final response. Does the response identify what was requested? Can its meaning be understood without guessing which intermediate quantity the student intended? This does not simulate every marking rule, because real markers may consider working. It tests the communication quality of the ending.
For example, present the question “How much more did the second group collect?” beside the answer “They collected 86.” The number may be correct for the second group, but the difference is not stated. Present “Give the coordinates of the turning point” beside “x = 2.” One coordinate has been found, not the complete requested object. Present “State whether the evidence supports the claim” beside a correct list of observations. The conclusion is still missing.
Then reveal the working. In each case, ask whether the missing result is already available somewhere on the page. If it is, the student needs to select and communicate it. If it is not, another reasoning step remains. This distinguishes copying the wrong result from stopping the solution early.
Reverse the exercise by showing a clear final response with incomplete working. A correct conclusion does not automatically prove that all required reasoning has been demonstrated. Where the task asks for justification, the route matters too. The purpose is to keep both obligations visible: produce an appropriate result and provide the evidence the assessment requires.
This rehearsal works best on a small sample. Do not convert every answer into a lengthy discussion. Once the student can recognise the difference, ask them to perform the same check silently on their own work. The eventual habit is brief: read the question beside the answer and confirm that they match.
21. Trial the finishing routine without changing everything else
Choose one known completion problem for a short teaching trial. Perhaps the learner repeatedly reports an intermediate quantity. Keep the normal lesson structure and add only the quantity-label check. Use several fresh questions in which more than one meaningful number appears, then examine whether the correct quantity reaches the final response.
Do not simultaneously introduce a new solution method, a new worksheet format, a stricter timer and a new marking standard. If performance changes under all those conditions, it becomes difficult to know what helped. A small trial makes the result easier to interpret, although it is still ordinary teaching evidence rather than a controlled causal experiment.
Record both benefit and cost. Did the check prevent the known error? Did the learner start labelling every tiny arithmetic operation and become unnecessarily slow? Did the routine help on one context but disappear when the wording changed? Those observations determine the next adjustment.
A second session should include the target demand among other question types. Do not announce that every item is a final-answer trap. The learner must notice when the check is relevant. Include a question whose first calculated value is already the final answer, because a good routine should also permit stopping.
When the routine works only with a reminder, keep practising independent initiation. When it works independently but takes too long, simplify its wording. When it fails because the student misunderstands the quantity, return to conceptual teaching. When it works across fresh material at acceptable cost, retain it as light maintenance rather than continuing an intensive repair block.
The outcome to seek is not perfect use of the teacher’s script. It is a more reliable finished answer. Students can use different internal wording and still achieve that outcome. A teaching routine is successful when its useful function remains after its visible scaffolding fades.
22. What students, parents and teachers should say next
Replace “You did all the hard work and threw away the mark” with a question: “Which final decision was still unfinished?” The first sentence describes frustration. The second opens a repair.
A student can ask, “Was my model wrong, my calculation wrong, or my finished response wrong?” A parent can ask, “What small change will let you notice this next time?” A teacher can ask, “Can you finish a changed example without me naming the missing step?” Each question moves from blame toward evidence.
Do not make the answer longer merely to make the work look serious. An exact expression, one valid unit conversion or one clearly selected solution may complete the task. In writing, a precise relationship may do more than another paragraph of related facts. Completion is about sufficiency, not volume.
The final mark is often protected by a return: return to the original question, return to the meaning of the number, return to the conditions, and return to the place where the answer must be submitted. The learner does not need to distrust everything they have done. They need to check whether what they have done has reached the requested destination.
Correct working is valuable. Clear, valid completion makes that working usable to the reader and assessable under the relevant rules. Teach the finish as part of the task, then let the student move on.
Sources and further reading
AQA: GCSE Statistics command words supplies a qualification-specific example of answer-form and placement requirements. OCR: How accurate is a mathematical answer? explains precision and exactness. Neither is treated here as a universal marking rule.
Continue through the Complete Examination Craft Index for paper-performance guidance, or the Learning Runtime Hub when the source of the difficulty remains unclear.
