A literature review groups existing research by theme, synthesizes it rather than summarizing source by source, and ends by naming the gap your project fills. How to write a literature review comes down to seven steps: focus the question, search systematically, screen the results, extract sources into a synthesis matrix, find the themes, outline by theme, then write and revise. This guide also covers types, format and length.
how to write a literature review
What Is a Literature Review, and How to Write One: Steps, Format and Types
On this page
- What is a literature review?
- Types of literature review
- How to write a literature review in seven steps
- Synthesis vs summary: the paragraph that earns the grade
- Literature review format and layout
- How long a literature review should be
- How to start and how to end a literature review
- See a complete example
- When the review is due before the reading is done
What is a literature review?
What is a literature review, in plain terms? It is a structured account of what researchers already know about a question, organized so a reader can see where the field agrees, where it disagrees, and what it has not yet answered. It is not a list of summaries; a stack of one-paragraph-per-source write-ups is an annotated bibliography, a related but different document.
Every literature review, whatever its length, is doing the same underlying job: turning a pile of separate findings into one coherent account of a field. A reader should be able to finish it and answer three questions without going back to the sources themselves: what does the field currently agree on, where does it disagree, and what has nobody tested yet. A review that cannot answer those three questions has summarized its sources without actually reviewing them.
What a literature review is for
The purpose of a literature review is to justify a project before it starts, not to prove you did the reading. A strong review shows a reader that your research question is not already answered, that your chosen method fits how the field has studied the question so far, and that your project builds on specific findings rather than starting from nothing. Skipping this step is the fastest way to propose a study that already exists, and committees catch that gap in the proposal stage far more often than students expect.
A literature review also does quieter work that rarely gets named directly: it teaches you the field's own vocabulary, shows you which researchers keep citing each other, and often surfaces the exact method or instrument you will end up using yourself. Treating the review as a box to check before the "real" project starts wastes most of that value; treating it as the first real piece of research keeps it.
Literature review meaning in a research paper vs a dissertation
The core moves stay the same, but the scale and the placement change. Inside a shorter paper, a literature review inside a research paper is usually a background section a few paragraphs long, built from a smaller source set and folded into the introduction rather than standing alone. Inside a dissertation or thesis, the literature review is its own chapter, built from a much larger source set, and it earns its length because where Chapter 2 sits in the thesis is directly after the introduction and directly before the methodology chapter it has to justify.
Types of literature review
Most instructors mean a narrative review by default, but knowing the other types helps you recognize which one your assignment, or your committee, actually expects.
| Type | What it does | When it's used |
|---|---|---|
| Narrative | Thematic synthesis without a fixed search protocol | Course papers, dissertation introductions, broad topic overviews |
| Systematic literature review | Follows a protocol set in advance, an exhaustive search and explicit inclusion and exclusion criteria | Health sciences, evidence-based fields, where reproducibility matters |
| Scoping | Maps the breadth of a field and its gaps, without appraising each source's quality | A field too broad or too new for a fixed systematic method |
| Integrative | Combines qualitative and quantitative sources into one synthesis | Social sciences and nursing, building a new conceptual framework |
A systematic literature review is the most demanding of the four: it follows a protocol written before the search and often registered in advance (for example, in PROSPERO), records every search term and database, and reports its method to a reporting guideline such as PRISMA, including a flow diagram of how many records were found, screened and excluded at each stage. Most course and dissertation-chapter reviews are narrative, not systematic, even when they are thorough; a narrative review can still search widely and screen carefully, it simply does not commit to reporting every step of that process the way a true systematic review does. Ask your instructor or your committee which type is actually expected before you start, since a narrative review presented as systematic, or the reverse, reads as a structural mismatch a careful reader notices immediately.
How to write a literature review in seven steps
How do you do a literature review in practice? Seven steps, done roughly in order, though reading and refining the question usually loop back on each other more than once. Many writers build an annotated bibliography, the step before the review, first: a short summary and evaluation of each source on its own, before any synthesis begins.
How to write a literature review in seven steps
Form a focused research question narrow enough that a database search returns a manageable, relevant set of results.
Search systematically across two or three databases suited to your field, recording your search terms as you go.
Screen every result against clear inclusion and exclusion criteria, and note why you excluded a source, not just that you did.
Read and extract each included source into a synthesis matrix, sources as rows and themes as columns, not a separate summary paragraph per source.
Find the themes: where sources agree, where they disagree, and what no source in your set has yet answered.
Outline the review by theme, never by source, so each section becomes an argument rather than a list.
Write the draft, then revise it against the synthesis-vs-summary test in the next section.
Step 4's synthesis matrix is the single most useful habit in that list. Build it as a simple grid, one row per source and one column per recurring theme, and note in each cell what that source actually says about that theme, not just whether it mentions it.
| Source | Method | Theme: access | Theme: reported stress |
|---|---|---|---|
| Smith (2019) | Field observation | 20-minute park visit | Lower, self-reported |
| Chen & Okafor (2021) | Survey, 300 office workers | Window view at work | Lower reported burnout |
| Diallo (2022) | Interviews, one neighborhood | Little to no green space nearby | Frustration named directly |
Scroll the table sideways to see every column.
Read down a column, not across a row, and the themes your outline needs are usually already visible before you write a single sentence of the review itself. A spreadsheet works as well as any specialized tool for this step; the value is in the habit of extracting by theme, not in the software that holds it, and a matrix you can scan in one sitting beats a folder of separate notes you have to reopen one at a time.
Synthesis vs summary: the paragraph that earns the grade
A source-by-source paragraph and a synthesized paragraph can cover the exact same three sources and read as two different levels of work.
Same three sources, summary vs synthesis (composite sample written by GradeDraft)
Before
Smith (2019) found that visiting a park for 20 minutes reduced self-reported stress in urban adults. Chen and Okafor (2021) surveyed 300 office workers and found that a view of greenery from a workspace window correlated with lower reported burnout. Diallo (2022) interviewed residents of a low-green-space neighborhood and reported that many described wanting more access to parks.
After
Across observational, survey and interview methods, exposure to green space consistently tracks with lower self-reported stress, whether that exposure is a 20-minute park visit (Smith, 2019), a window view at work (Chen & Okafor, 2021), or its simple absence, which Diallo's (2022) interviewees named directly as a source of frustration. That the pattern holds across three different methods strengthens the case that the association is not an artifact of any single one of them.
What changed
- Three separate source sentences merged into one claim about a pattern
- The variety of methods becomes evidence for the pattern's strength, not just a list of studies
- A source-by-source list becomes an argument a reader can actually use
Notice that the "after" version cites the same three sources, in the same parentheses style, and adds no new evidence. The only thing that changed is which sentence does the work: the sources, or the claim the sources are made to support together.
Literature review format and layout
The format of a literature review and its layout both follow from one decision made early: how will the sections be organized?
| Structure | Organizes by | Best when |
|---|---|---|
| Thematic | Themes or concepts that cut across sources | Sources cluster into a small number of clear ideas (the most common choice) |
| Chronological | Time, oldest sources to newest | The field's thinking visibly changed over time |
| Methodological | The method each source used | Comparing what different methods find on the same question |
Most reviews, at every level, default to a thematic structure, because it is the structure that best supports synthesis: a chronological review tends to drift back toward summary, one source per era, unless the writer works to resist it. A methodological structure is the least common of the three outside of fields where competing methods are themselves the story worth telling, such as a field split between lab and field studies of the same question.
Layout follows structure once it is chosen: each theme, era, or method becomes its own labeled subsection with a short topic sentence stating what that section will show, followed by the synthesized paragraphs themselves, and closing with a sentence connecting that section back to the review's overall gap. A layout that never states each subsection's point up front forces a reader to reconstruct your argument instead of following it. Once the literature review states its gap, the methodology chapter that follows is where that gap turns into a specific research design.
How long a literature review should be
How long should a literature review be? There is no fixed rule: length depends on the level of the document it belongs to and how much has already been published on the question.
| Level | Typical length |
|---|---|
| Course research paper | 2 to 4 pages, usually folded into a background section |
| Master's thesis chapter | 15 to 25 pages |
| Doctoral dissertation chapter | 25 to 40-plus pages |
Treat every number in that table as a starting expectation, not a target to hit exactly; a narrow, well-sourced field can produce a strong review shorter than the range above, and a sprawling one can reasonably run longer. Padding a thin field's review to hit a page count is more visible to a reader than a short, well-synthesized one, since padding almost always shows up as repeated summary rather than new synthesis. Your own program's handbook or your committee's guidance always overrides a general range like this one.
How to start and how to end a literature review
How to start a literature review comes down to one sentence that states three boundaries at once: the date range covered, the population or field the sources address, and the outcome or construct being measured. That single scope sentence tells a reader what is in the review and, just as usefully, what is deliberately left out.
Ending works the same way in reverse. A strong closing paragraph does not restate the sources; it names the specific gap, ideally as a missing comparison or an untested case, and states plainly that this is the gap the current project addresses. A vague call for "more research" is the most common way a review ends without actually earning its conclusion.
Between the opening scope sentence and the closing gap statement, each themed section should also open with its own one-sentence claim, then support it, the same rule that governs a strong paragraph anywhere in academic writing. A reader who reads only the first sentence of every section should still walk away with an accurate, if thin, summary of your entire argument.
See a complete example
Every move covered above, the scope sentence, synthesis across multiple sources, a stated disagreement, and a gap statement, is easier to spot inside a finished review than in a list of rules. A complete APA 7 literature review, annotated marks all four with margin notes and shows the synthesis matrix behind it. Read it before you draft your own section headings.
When the review is due before the reading is done
A literature review is often the slowest chapter to write, since the reading has to happen before the synthesis can start. Screening and extraction usually take longer than planned, and that is where a deadline starts to slip. When the deadline is the real constraint, GradeDraft's dissertation help by milestone covers every stage from a single chapter to a full manuscript. A standalone literature review written for you covers this chapter on its own, built as a synthesis matrix rather than a source-by-source summary, from your topic and your required source count. When the review is one chapter among several still due, Chapter 2 written as part of the dissertation covers it alongside the chapters around it.
Send your topic, your required source count, and your citation style. A confirmed writer and a fixed price come back before anything is invoiced.
Get my literature review writtenFAQ
Literature review questions
Replies within 2 hours, 8 am–11 pm ET, 7 days a week
01What does a literature review look like?
A finished literature review reads as a small number of themed sections, each one synthesizing several sources into a claim about what the field currently supports, disagrees on, or has not yet tested. It opens with a scope sentence naming the date range, population and outcome covered, and closes with a stated gap, and it rarely, if ever, gives each source its own separate paragraph the way an annotated bibliography does.
02How many sources should a literature review have?
Enough to reach saturation: the point where one more source stops adding a new theme and only adds another citation to a theme you have already covered. A course paper might synthesize 8 to 12 sources; a dissertation chapter often needs several times that, sometimes fifty or more depending on how established the field is. Your program's guidance, where one exists, always sets the actual floor over any general number like these.
03Is a literature review the same as an annotated bibliography?
No. An annotated bibliography gives each source its own separate entry and summary, one at a time. A literature review groups sources by theme and synthesizes several of them inside a single sentence or paragraph, then states what the existing research has not yet answered. Reading like a list, even a well-written, accurate list, is the clearest sign a draft is still an annotated bibliography and not yet a review.
04How do you do a literature review?
Form a focused question, search systematically across databases suited to your field, screen what you find against clear inclusion criteria, extract each source into a synthesis matrix, find the themes across sources, outline by theme rather than by source, then write and revise against the synthesis test above. The order rarely runs perfectly straight; an early read often sends you back to refine the question or the search terms before you move on to extraction.
05What is the purpose of a literature review?
The purpose of a literature review is to show that your research question is not already answered, that your planned method fits how the field has studied the question so far, and that your project builds on specific, named findings rather than starting from an unexamined assumption. A committee reads it to judge whether the rest of the project is justified before any data is collected or a single chapter past it is written.