What a conclusion does that an abstract does not
An abstract and a conclusion contain overlapping facts and have opposite functions. The abstract is a proportional miniature of the paper, written for someone deciding whether to read it, and it covers all four moves including the method. The conclusion is written for someone who has just finished reading, and it covers only the answer and its consequences.
This is why a conclusion that recaps the method reads as redundant. The reader has been through the method; repeating it uses the last paragraph of the paper to tell them something they already know. Everything in the conclusion should look forward from the finding rather than back at the work.
The difference shows up in the verbs. Abstracts report — “participants completed”, “accuracy improved”. Conclusions assert and project — “the effect depends on timing”, “courses that batch feedback should not expect the same benefit”. If your conclusion is written entirely in reporting verbs, it is an abstract in the wrong place.
The one-sentence answer
Almost every strong conclusion contains a single sentence that a reader could carry away and repeat correctly. Writing it is harder than it looks, because it has to be specific enough to be informative and hedged enough to be defensible. Draft it separately, outside the paragraph, and then build the section around it.
Weak: “This study contributes to our understanding of feedback in higher education.” That could be attached to any paper on the topic. Stronger: “Feedback improves accuracy when it arrives in the same session, and the benefit largely disappears at a one-week delay.” The second commits to something, which is why it is worth reading.
If the sentence will not come, the usual cause is that the paper answers a different question from the one it asked, or answers several partially. That is a structural problem worth finding before submission rather than after review, and the conclusion is where it becomes visible. The repair is usually to narrow the question rather than to widen the paper.
From contribution to consequence
A contribution statement says what the field gains. It works best when it is stated in the vocabulary of the debate you entered in the introduction, so the reader can see the loop close. If your introduction framed a contradiction between two accounts, the conclusion should say which account your evidence favors and how strongly.
Consequence is the step most conclusions omit. Having stated what is now known, say what follows: which practice should change, which model needs revising, which measurement should be preferred, which decision becomes better informed. Attach an actor where you can, because “implications for practice” with no practitioner named tends to mean nothing in particular.
Keep the size of the claim proportional to the design, and keep it consistent with the hedging you used in the discussion. A conclusion that is noticeably bolder than the section preceding it is one of the fastest ways to lose a reviewer’s trust, and it is easy to do accidentally when the conclusion is written under deadline. Reading the discussion and the conclusion one after the other is the quickest way to catch the drift.
Conclusions in theses and in articles
A journal article conclusion is short — often a single paragraph, sometimes folded into the end of the discussion when the journal does not use a separate heading. Its whole job is the answer, the contribution, and the consequence. Length here is not a virtue.
A thesis conclusion is a chapter, and it does more work: it draws together findings that were reported across several empirical chapters, states the overall contribution, addresses the limitations of the project as a whole rather than of one study, and sets out a research agenda. It still should not summarize each chapter in turn, which is the version examiners see most often. What they are looking for is the argument only the completed thesis could make.
The organizing move for a thesis conclusion is to answer the overarching research question first, then show how each empirical chapter contributed to that answer. Written that way, the chapter argues; written the other way round, it becomes a table of contents in prose. One paragraph per empirical chapter, each framed as a contribution to the overall answer, is usually enough.
Endings that weaken a paper
Three endings recur and all of them cost something. The first is the retreat: closing on limitations, so the last thing a reader takes away is what the study could not do. The second is the ritual call for more research, which every paper could make and which therefore says nothing. The third is the grand claim that the study cannot support, which sends a careful reader back to check whether the rest of the paper was honest.
A useful check is to read the final sentence alone and ask whether it is true, specific, and yours. If it would fit at the end of any paper in your field, replace it. If it claims more than your discussion argued, pull it back to match.
Rhetorical flourishes rarely help in academic prose. A quotation, a rhetorical question, or a return to an opening anecdote can work in essays and in some humanities writing, but in an empirical paper the ending that lands is a plain statement of what is now known and what should be done about it. If you want a memorable last line, make it the plainest sentence in the paper rather than the most decorated.
