The Synthesis Closer: connect the sources, land the gap
Most failing review endings share one disease: the sources stay strangers. "Study A found this. Study B found that. Study C found the other." Three sentences, three islands. A closing paragraph earns its place only when the sources start talking to each other — one supports, one pushes back, one carries the argument somewhere new — and the paragraph ends by naming what the whole conversation never got around to.
That is what the tool below assembles. Write one short summary per source — plain language, no citation needed for the draft — and tag its stance: does it support the main line, contest it, or extend it? Add a gap sentence if you have one. The Closer then builds a draft paragraph that connects the stances, and a meter tells you whether you have drifted into list territory. Everything stays on this device; the draft saves locally so a reload brings it back.
Synthesis Closer — sources in, connected draft out
3-4 source summaries + stance tags → connected paragraph · list-vs-synthesis meter · draft saved on this device
Scoped beats grand: name the measure, group or setting nobody tested.
Your connected draft will appear here — two or more filled sources, stances tagged.
How the meter reads your draft: three ingredients, roughly a third each. Stance mix — if every source got the same tag, the paragraph has no friction to synthesize, which is how laundry lists happen. Connective logic — the assembled lines must join sources with while, contests, extends, agrees — the grammar of relationship, not of inventory. The gap line — a named hole turns the paragraph from a recap into a setup. All three on means a closer that does its three jobs.
A word on tagging, because this is where most students stall. Supports means the source's findings land on the same side as your review's core claim — it can qualify, scope or extend the numbers and still count as support. Contests is not "mentions a limitation"; it means the source argues against a finding or assumption another source relies on, which is why you need at least one if the paragraph is to argue anything. Extends is the source that carries the line somewhere new — a new population, a new method, a new setting — without disputing the core. When two tags seem to fit, pick the one a classmate would guess from your one-line summary, because that is the stance your paragraph will assert. Mis-tags produce mis-funnels: a review with two phantom contests will promise a debate the sources never had.
What the ending must do: three jobs
Strip away the advice and a literature review conclusion has exactly three jobs. Everything else — citations, hedging, signposting — is either serving them or smothering them.
Job 1: compress the field back into one argument
The body of your review spread the sources across themes; the ending fuses them again. The test is simple: could a reader who only saw this paragraph say what the field currently believes, and where it splits? For an urban-heat review, the compressed version sounds like: surface materials drive daytime street temperatures, canopy coverage helps but cannot carry the whole burden alone. Two clauses, one field. If your version needs a paragraph per source, the compression has not happened yet.
Job 2: name the gap — specifically
"More research is needed" names nothing; it is a shrug with a citation habit. A gap statement points at a measure, a population or a setting the reviewed sources never touched. The heat review might land on: whether night-time heat moves indoors, where the corridor surveys stop measuring. That sentence does three things at once — it admits the field is incomplete, shows you read carefully enough to find the seam, and hands your project its reason to exist. One sentence, specific to the studies above it.
Job 3: set up your study — the funnel in reverse
The closing sentence turns from the field toward you. It does not need to restate your whole research question; it needs to make the reader expect one. For a school-lunch review, the bridge is one clause: what no pilot program has tested is whether uptake survives when the menu rotates weekly — which is the question this project takes into two district cafeterias. Field, seam, study. When all three sentences exist, the ending reads inevitable rather than bolted on.
The three jobs also explain why the Synthesis Closer's meter is built the way it is. Stance mix is Job 1 — agreement without dissent compresses into mush. Connective logic is the grammar all three jobs run on. The gap line is Jobs 2 and 3 fused: a gap named specifically is already half a handoff. If your draft scores low, the fix is never "write more" — it is one of the three ingredients going missing.
The no-new-sources rule
Halfway through drafting the ending, you will meet it: the perfect study, published this year, exactly on your gap — and never discussed in your body. The rule is blunt. It does not go in the conclusion. A source mentioned for the first time in the final paragraph has no section beneath it: nothing summarized it, nothing weighed its method, nothing connected it to the themes. It arrives as an assertion, and assertions in a conclusion read as padding.
What to do instead: slot it into the body where it belongs, review it like every other source, and let its point surface through the synthesis you already wrote. If it genuinely reshapes the gap — say a school-lunch pilot already tested rotating menus and failed — then the gap statement changes, honestly, and the conclusion says what the field still has not done after that study. The ending may compress anything the body covered. It may not smuggle.
Conclusion, abstract or executive summary?
Three compressed texts, three different masters. Students blur them because all three are short and all three summarize — but they answer different readers with different clocks. The lit review conclusion closes the literature; the abstract sells the whole paper in a database; the executive summary briefs a decision-maker who may never read past it. The table sorts them.
| Dimension | Literature review conclusion | Abstract | Executive summary |
|---|---|---|---|
| Length | One paragraph, roughly 100–200 words | 150–300 words for the entire paper | One page or less, standalone |
| Tense | Past or present perfect for sources; present for the gap and your study | Past for what was done, present for what it means | Present for findings, future for recommendations |
| Audience question answered | "So what is still missing — and why this study?" | "Is this paper worth retrieving and reading?" | "What should the organization decide?" |
| Placement | Last paragraph of the review, inside the paper | Front matter, before the introduction, with keywords | Front of a report, written last, read first |
The mix-up that costs the most marks: writing the review's ending like an abstract. An abstract may announce the paper's plan; a review conclusion must first pay the field its due — the compression and the gap — before the handoff. If your ending would still make sense lifted out of the paper and pasted above a database record, it is doing the wrong job.
Order of writing helps too. The abstract is written last but sits first; the executive summary is written last and read first; the review conclusion is written last and read in place — which means it can lean on everything above it. Use that license. Phrases like "the corridor surveys above" or "the three pilots reviewed here" let the ending compress whole sections into two words, a move no abstract can afford because its reader has not read the sections yet.
Two model closers: one connects, one lists
Both closers below are invented for practice, built on the same three-source set. The first connects the stances and lands the gap; the second recites the sources and hopes. Read each, then tap Why it works — or fails.
Model 1 · connected synthesis · urban heat islands
"Across these studies, one argument holds: what a street is paved with decides how hot its afternoons get. The corridor surveys and the rooftop field trials agree on that core, and the simulation work extends it block by block. The disagreement is narrower than it looks — canopy coverage helps, but only where surfaces have already been fixed. What no source in this review measures is the night: whether heat built up on the asphalt all day then moves indoors after dark. That untested seam is where this project begins."
Every sentence relates sources instead of listing them: two groups agree, one extends, the dissenter gets folded in with a scope. The gap is anchored to the exact measurement the sources stop at, so the handoff to the study feels earned. Notice there is no per-source recap — the body already did that work, and this paragraph trusts it.
Model 2 · the list-style stall · school lunch programs
"One pilot program found that salad bars increased vegetable uptake. A second study found that longer lunch periods helped. Another study found that chef-designed menus helped too. In conclusion, more research is needed on school lunches."
Three findings, three islands — no sentence says how uptake, time and menus relate, so the paragraph is a recap with a stamp. The final line is the emptiest move in review writing: it names no measure, no setting, no seam. Delete the sources and the paragraph survives, which is the tell. The gap sentence could name, for example, whether uptake survives weekly menu rotation — but the list version never gets there.
Four endings that fail on contact
Four shapes account for most of the damage. Each card states the failure the way a marker would see it; flip for the repair.
The second related-work section
Tap to flip
The ending restarts the survey — new clusters, new comparisons, body two. A conclusion compresses what the review already argued; it never widens the lens. If a theme deserves a full section, promote it into the body and let the ending spend its 150 words on fusion, seam and handoff.
The source laundry list
Tap to flip
"A found X. B found Y. C found Z." — findings in citation order, zero relationships. The fix is stance grammar: mark who supports, contests, extends, then join them with while, although, together. If any two sources could swap places without breaking a sentence, nothing is connected yet.
"More research is needed" — no gap named
Tap to flip
Every review ever written could end this way, which is why it says nothing. Name the hole: the measure, group or setting the sources never reached. Compare "more research is needed on heat" with "no corridor survey measures where the heat goes at night" — the second one recruits your study.
The last-minute source dump
Tap to flip
Two fresh citations in the closing lines, never reviewed. The no-new-sources rule applies: an unreviewed source cannot be synthesized, only name-dropped. Move it into the body and review it properly, or drop it. The conclusion may compress the field the body covered — never expand it.
The funnel: field, seam, study
Think of the review's shape as a funnel running the wrong way for most of its length — wide at the top, narrowing theme by theme — and then opening again at the very end, where your study walks in. The ending manages that final turn. Four moves, roughly one sentence each:
Field
One clause for what the reviewed sources collectively establish — the compressed claim from Job 1.
Debate
Where the stances split, in one sentence — supports against contests, with the extenders building on whichever side holds.
Seam
The gap sentence: the specific measure, group or setting nobody in the review touched. Scoped, not grand.
Study
The handoff clause — "that is where this project begins" — pointing at your question without restating your whole design.
Run the urban-heat example through it: field — paved surfaces drive afternoon street heat. Debate — canopy coverage helps, but only after surfaces are fixed. Seam — no source measures where that heat goes at night. Study — a night-time indoor sensor log in two corridors. Four sentences, and the review is over without ever saying "in conclusion". For the school-lunch review the same funnel ends on uptake: field — programs change what students eat; debate — the levers argued over are time, menus and placement; seam — none of the pilots tested weekly menu rotation; study — a term-long rotation trial in one district.
The funnel's most common failure is skipping the debate step. Without a stated disagreement, the seam arrives from nowhere: the reader has watched a field agree, then suddenly must believe something is missing. Ground every gap in the fight above it — the untested night follows because the day is settled, and the rotation question follows because the pilots each tested one fixed menu. A gap the funnel earned reads as rigor; the same gap dropped in cold reads as a wish.
Ending drafted, deadline still circling?
When the review is mapped and the gap is named but the weeks are gone, a deadline-bound writer can take the paper on and hand back a draft whose closing paragraph you can hold against everything on this page — synthesis, seam and handoff included.
Quiz: five questions on closing the review
Five questions, one correct answer each; a wrong pick reveals the right one. Five for five means the ending rules stuck.
1. The last paragraph of a literature review exists to…
2. "Study A found X. Study B found Y. Study C found Z." What is broken?
3. Which closing line earns its place?
4. While drafting the ending you find a perfect new study on your gap. What now?
5. A literature review conclusion differs from an abstract because the conclusion…
Literature review conclusion: frequently asked questions
How long should the conclusion of a literature review be?
One tight paragraph for a course paper, two at most for a thesis chapter. It has three sentences worth of work — compress the themes, name the gap, hand off to your study — and every sentence past that job starts re-running the body. If the ending stretches past roughly two hundred words, the likely cause is a theme that was never synthesized in its own section, so the conclusion is quietly doing the body's job.
Can I mention sources in the conclusion that I never discussed in the body?
No. The no-new-sources rule exists because a conclusion has no room to summarize, let alone synthesize, a source the reader never met. A study dropped into the last paragraph cannot be weighed, compared or contested — it can only be asserted. If the source truly matters, it goes in the body where it belongs, and the ending absorbs its point through the synthesis you already wrote.
How do I name the gap without overclaiming?
Anchor the gap to the sources you actually read, and scope it to a measure, a population or a setting. Compare a vague opener like "more research is needed" with a scoped one: "whether night-time heat moves indoors is untested in the corridor surveys above." The second version names what is missing and who failed to test it, which is why it sounds confident instead of needy.
What tense should a literature review conclusion use?
Past tense or present perfect for what the sources did — agreed, contested, extended — because studies are finished work. Present tense is reserved for field-level claims that still hold and for your own study's plan: "this review leaves open," "this project will measure." Tense wobble usually signals a deeper wobble between reporting the sources and starting to speak for yourself, so fix the roles first.
Does the conclusion of a literature review need citations?
Usually no, and that absence is the point. The ending compresses what the cited sections established, so it refers to themes and stances rather than individual studies. Citing fresh authors there trips the no-new-sources rule, and re-citing old ones pads the paragraph. The exception: a phrase like "the corridor surveys" may echo a study group already named in the body.
Which failing ending is your draft closest to?