Did you know that the most confident sentence in a student essay can be the one the essay has not actually earned? It often appears at the very end. After several sensible paragraphs, a writer wants a memorable finish and suddenly makes a new claim that no earlier evidence addresses.
An essay conclusion should bring the argument together, not quietly start another argument. To repair a conclusion, underline its central claim, trace that claim back to the reasons and evidence already developed, and either revise the ending to match the demonstrated reasoning or move the new point into the body and support it properly. This is not a ban on original insight: the ending should synthesise what the essay has established.
Purdue University’s OWL guidance on expository essays distinguishes effective synthesis from merely restating a thesis and advises against introducing unsupported new information in the conclusion. The method below applies that advice to ordinary classroom writing. All student writing and numbers in the examples are invented for teaching, not taken from real pupils.
The problem hidden inside a good-sounding ending
Consider a short essay on the question: Should a school pilot a voluntary lunchtime reading corner? A learner may describe the benefit of a quieter place, discuss how to make books accessible, and recommend trying the idea for a month. Then the ending declares, “This programme will raise every student’s examination marks.”
That sentence sounds decisive. Yet it introduces a much stronger outcome than any of the body paragraphs established. The writer has moved from access to reading opportunities to a promise about examination results. Those are different claims. There may be reasons to study their relationship, but the essay has not supplied the evidence required for a guarantee.
The error is not poor punctuation. It is a mismatch between the scope of the conclusion and the scope of the supporting argument.
Distinguish synthesis, summary and a new argument
A summary reminds readers of ideas already discussed. A synthesis shows how those ideas fit together and what narrower judgement follows from them. A new argument introduces an additional factual assertion, cause, comparison or recommendation that earlier paragraphs have not established.
The Purdue OWL resource on synthesising sources explains the difference between restating information and connecting evidence to reach an insight. An ending can therefore add a thoughtful interpretation without smuggling in an unsupported fact.
Ask a simple question: Could a reader point to the earlier paragraph that gives the reason for this conclusion? If not, the conclusion may be overreaching. This test is especially valuable when the final sentence uses words such as always, everyone, proves, will certainly or must, because these make a claim harder to justify.
Worked example: a classroom reading-corner proposal
Imagine a wholly fictional class project involving 24 participating students. During a four-week voluntary reading-corner pilot, students recorded whether they borrowed a book from a class shelf at least once. The records show 15 of 24 did so, and 9 of 24 did not. The project did not measure reading comprehension, test marks, home reading or long-term behaviour. Assume the counts are complete and each student is counted once.
The fraction who borrowed at least once is 15 ÷ 24 = 0.625, or 62.5% of participants. That calculation is valid. But the number measures borrowing in this pilot, not academic improvement. It also does not represent every student in the school.
Suppose the essay’s introduction says, “A small reading corner could make borrowing books easier for students who choose to use it.” Its body describes the shelf, the voluntary schedule, the 15 borrowers, and the need to keep access manageable. So far, the reasoning addresses feasibility and observed use.
The unearned conclusion
“Clearly, the reading corner has increased everyone’s English ability. Because 62.5% borrowed books, the school should immediately guarantee that examination results will rise.”
The mistakes are precise:
- Borrowing is not the same measurement as English ability.
- 62.5% is not everyone. Nine participants did not borrow a book during the pilot.
- No before-and-after comparison of English ability was made.
- No examination outcomes were recorded.
- The recommendation escalates from a small voluntary pilot to a promised school-wide result.
A justified revision
“In this four-week pilot, 15 of 24 participating students borrowed at least one book from the reading corner. That shows the shelf was used by a substantial share of this small group, although borrowing alone does not tell us whether reading skill improved. A sensible next step is to continue a manageable trial and investigate whether the corner supports regular reading.”
This version is less dramatic but stronger. It uses the actual observation, identifies the limit, and proposes a proportionate next step. The writer makes a new judgement from existing reasons, not a new factual claim disguised as a conclusion.
The three-line audit
A learner can inspect a conclusion with three short lines written in the margin:
- Claim: What exactly am I asserting in this sentence?
- Support: Which paragraph, source or worked result supports it?
- Limit: What have I not measured, explained or established?
For the fictional reading-corner essay, the claim “some participating students used the shelf” points to the borrowing register. The claim “reading ability improved for everyone” points nowhere in the evidence. The limit is that borrowing frequency cannot by itself establish learning gains.
This audit works better than a vague instruction to “make the conclusion stronger”. It gives the learner a test they can perform and a specific way to change the text.
When the new idea really matters
Sometimes an ending reveals an excellent new idea. Do not automatically delete it. Ask whether the essay should be expanded.
If the writer wants to argue that regular reading may support comprehension, the argument needs suitable earlier development: a clear definition of the outcome, relevant evidence, careful source attribution, possible alternative explanations, and wording matched to what is known. The writer can then return to the conclusion and synthesise this expanded discussion.
But adding one impressive sentence at the end does not replace that work. The middle must carry the argument’s weight.
If the assignment is short, the better choice may be to remove the ambitious final claim and keep the original, defensible topic. The revision depends on the task’s purpose and length.
An independent diagnostic for the learner
Give a student these three endings for a report about a fictional class garden, where the body only documented plant-height measurements over two weeks:
- “Our measurements show the measured seedlings grew between 2 and 5 cm during these two weeks.”
- “The seedlings must have become healthier because every plant in the world benefits from this fertiliser.”
- “The observed growth supports a cautious account of what happened in this class garden, but not a claim about every type of plant.”
The first describes a reported observation if the measurements indeed support it. The second adds a sweeping unsupported mechanism and universal claim. The third connects the evidence to a limitation. Ask the learner to explain why, not merely identify the “correct” sentence. Then remove the prompts and test the skill with a new subject.
A teacher can distinguish a comprehension difficulty (“I cannot find the supporting paragraph”), a scope problem (“I changed some into all”) and a writing problem (“I repeated the opening but did not bring the reasons together”). Each needs a different correction.
Guided practice in three passes
Pass one: highlight without rewriting
Read the essay’s first and last paragraphs together. Circle the thesis or central question and underline every new promise, cause or quantitative claim in the conclusion. Use a different colour for each main body reason. The student should locate support before changing wording.
Pass two: make the conclusion match what was demonstrated
Rewrite the ending in two or three purposeful moves: return to the essay’s question, connect the most important supported reasons, and finish with a measured judgement or next step. A conclusion need not use a stock phrase such as “In conclusion” if its function is already clear.
Pass three: test on an unfamiliar task
Give a short history explanation with a new claim in the final sentence, or a science report that claims a universal law from three measurements. Let the student apply the same claim-support-limit test independently. This is where a writing skill becomes a transferable reasoning skill.
A useful teacher-and-parent routine
For a teacher, feedback such as “unsupported claim in final sentence; which earlier paragraph establishes it?” is more actionable than “weak conclusion”. Ask pupils to attach a small support tag—paragraph 2, data table, cited source—to each important closing assertion. Then let them remove the tags before submitting the final draft.
A parent can read only the opening and the closing first, ask what changed in between, and invite the learner to show the evidence. The aim is not to supply a more impressive ending for the child. It is to help the child discover whether the ending follows logically from their own work.
Record observable progress with a few independent writing samples: fewer unsupported universal claims, clearer references to reasons, and a conclusion that can be traced to the body without a prompt. Do not guarantee marks based on one successful revision. Assignment rubrics and genre requirements still matter.
Frequently asked questions
Can a conclusion include a recommendation that was not word-for-word in the body?
Yes, when it is a reasonable, qualified judgement based on the argument already made. If it depends on a new fact or assumption, develop that support earlier or narrow the recommendation.
Does every conclusion need a statistic?
No. A literature interpretation may draw together textual details; a history explanation may weigh causes; a reflective piece may articulate a learning insight. The test is logical support, not the presence of numbers.
Should a conclusion simply repeat the introduction?
No. Repetition without synthesis misses a chance to show what the evidence means. Revisit the opening question in light of the developed reasons.
Is a new question allowed in the final sentence?
Often, if it is clearly framed as a question for future inquiry rather than asserted as a proven finding. Its suitability depends on the assignment. A speculative question should not contradict the body’s supported conclusion.
What if the essay body itself contains weak evidence?
A polished ending cannot rescue unsupported reasoning. Repair the body, improve the quality of evidence and attribution, and only then rewrite the conclusion.
The deeper purpose: accuracy before flourish
A strong ending is not simply a powerful-sounding ending. It is the moment when the learner demonstrates control over the relationship between evidence, reasons and judgement. That habit matters in written arguments, but also in science conclusions, project proposals and everyday decisions.
For connected practice, read how to keep citations with claims after rearranging an essay and how to turn teacher feedback into a verifiable second draft. The eduKateSG Education Hub offers broader learning pathways, while the Secondary English tuition guide and the Clementi Mathematics teaching reference illustrate explanation-first guidance and independent checking.
The learner’s next assignment can begin with a small promise: write only what the evidence earns. It is a modest rule with wide-reaching value.