How to write a literature review — structure, steps and example (2026)

Most students who sit down to write a literature review end up producing something else entirely: a stack of paragraphs that each summarize one paper, glued together in the order the papers were read. That is not a review — it is a reading log. Knowing how to write a literature review means learning to do the opposite: to read across a dozen or more sources, group them by idea, weigh where they agree and clash, and use that map to justify your own study. This guide gives you a five-step method, a structure you can reuse for a bachelor’s or master’s thesis, a synthesis technique that kills the “list of summaries” problem, and an annotated example so you can see synthesis on the page.

What a literature review actually is (and what it is not)

A literature review is a structured argument built from existing research. Its job is to show what is already known about your problem, where scholars disagree, which methods have been used, and — crucially — what gap your work will fill. It is evidence that you have read enough to have something new to say.

It is not a book report, and it is not a warehouse where you store every source you found. Every paragraph should advance a point about the field, not about a single author.

Literature review vs annotated bibliography vs summary list

These three are constantly confused, and the confusion is exactly what produces weak chapters. The table below draws the line.

Format	Organized by	Voice	Purpose
Summary list	Source (one paper per paragraph)	The authors’	None — it just reports
Annotated bibliography	Source (alphabetical)	Yours, briefly	Catalogue + short evaluation of each item
Literature review	Theme or argument	Yours throughout	Synthesize the field and expose a gap

If you can move your paragraphs around without breaking anything, you have a summary list. In a real review, the order carries the argument, so paragraphs cannot be shuffled freely.

Narrative vs systematic review — which one you need

Before you write a word, decide which kind of review your degree requires. A narrative review tells a reasoned story about a topic, selecting the most relevant work — this is what most bachelor’s and many master’s theses use. A systematic review follows a pre-registered, reproducible search protocol and reports every step, and it is common in medicine, psychology and evidence-based fields.

Dimension	Narrative review	Systematic review
Search	Purposive, selective	Exhaustive, documented
Reproducible	Not fully	Yes, by design
Typical length	8–20 pages	20+ pages, often a study in itself
Best for	Undergraduate, most master’s	Medicine, meta-analysis, PhD
Bias risk	Higher (selection)	Lower (protocol-controlled)

Tip: If your supervisor has not specified, assume a narrative review — but stay transparent about how you searched. Even a narrative review benefits from naming your databases and keywords, because that is what separates a scholarly review from an opinion piece.

Step 1 — Search for sources systematically

A review is only as strong as the reading behind it. Random searching produces a lopsided review that misses foundational work, so treat searching as a method in its own right. This step pairs naturally with choosing a workable thesis topic, because a topic you cannot find sources for is a topic you cannot review.

Where to look

Start broad, then go deep. A practical order for 2026:

• Google Scholar — for scope, citation counts and the “Cited by” trail that leads you to newer work.
• Your university library’s databases — Scopus, Web of Science, JSTOR, EBSCO, or field-specific ones, usually free through your student login.
• DOI-indexed journals — a stable DOI is a signal you are citing a real, findable source rather than a broken link.
• Reference lists of the best two or three papers you find — “backward searching” surfaces the classics everyone builds on.

Keywords and boolean operators

Build a small keyword grid: your main concept, its synonyms, and the outcome you care about. Then combine them. Quotation marks force exact phrases ("remote work"), OR widens a search (teenager OR adolescent), and a minus sign excludes noise (motivation -workplace). Run each combination, log what you find, and keep a short search diary — databases used, terms, dates. That diary is what lets you later write one honest sentence about how the review was built.

Step 2 — Read with a synthesis matrix

Reading source by source and writing as you go is the single biggest cause of summary-style chapters. Instead, extract into a synthesis matrix: a table where rows are sources and columns are the themes, variables or findings you care about. Fill it in as you read.

Source	Definition of concept	Method	Key finding	Limitation
Kowalski (2021)	Broad, behavioral	Survey, n=200	Positive effect	Single country
Nguyen (2023)	Narrow, cognitive	Experiment	No effect	Small sample
Adeyemi (2024)	Behavioral	Meta-analysis	Effect, but moderated	Publication bias

Read the matrix down the columns, not across the rows. Reading down “Key finding” instantly tells you that two studies found an effect and one did not — that tension is a paragraph. The matrix converts a pile of PDFs into groups of claims, and groups of claims are exactly what you write about.

Step 3 — Organize the chapter

Once themes emerge from the matrix, choose an organizing principle. Do not organize by author.

• Thematic (most common): one section per sub-topic or debate. Best when the field splits into clear issues.
• Chronological: trace how thinking evolved over time. Use it only when the development is the point, not as a lazy default.
• Methodological: group by how studies were done (qualitative vs quantitative, lab vs field). Useful when methods drive the disagreements.

A typical structure moves from general to specific: an opening that frames the field, two to four thematic sections, and a closing section that states the gap. This mirrors the funnel logic you also use in a full bachelor’s thesis workflow, where the review narrows the reader toward your research question.

Step 4 — Write the synthesis, not summaries

Here is the move that defines a good review. A summary sentence starts with an author: “Kowalski (2021) found X.” A synthesis sentence starts with an idea and pulls authors in as evidence: “Behavioral definitions dominate the recent work (Kowalski, 2021; Adeyemi, 2024), yet experimental studies question the effect (Nguyen, 2023).”

Compare the two:

Summary (weak): Kowalski (2021) surveyed 200 people and found a positive effect. Nguyen (2023) ran an experiment and found no effect. Adeyemi (2024) did a meta-analysis and found a moderated effect.

Synthesis (strong): Survey-based studies report a clear positive effect (Kowalski, 2021), but this is not settled: an experimental design found none (Nguyen, 2023), and a 2024 meta-analysis reconciles the two by showing the effect holds only under specific conditions (Adeyemi, 2024). The disagreement appears methodological rather than substantive — which is precisely the gap this study addresses.

The strong version names authors, but the sentence is about the field. It also sets up the gap in its last line. Getting the in-text citations right across styles matters here; if you are unsure when a claim needs a citation, see the guide to citing sources in academic writing.

If you have your search done and your matrix filled but the synthesis itself is the wall you keep hitting, Smart-Edu’s academic writing tool can draft a thematically organized review with in-text citations from the sources you feed it, typically in 30–90 minutes — a fast first draft you then verify, edit in your own voice and correct against the originals. Treat its output as scaffolding to check, never as a finished chapter to submit.

Step 5 — Identify and name the research gap

Every review must end by earning your study. A gap is not “no one has studied this” (usually someone has). It is more precise: a population not tested, a method not applied, a contradiction not resolved, or a context — country, era, industry — not covered. The synthesis in Step 4 already pointed to one: the methodological disagreement nobody has settled.

State the gap in one or two explicit sentences and connect it directly to your research question. At master’s level this section carries more weight and often anticipates your theoretical framework — the master’s thesis with AI guide covers how to scale the review up for graduate work.

Source recency and common mistakes

As a rule, keep most sources within the last five to ten years, with exceptions for seminal work that the field still cites — a foundational theory from 1975 belongs even if it is old. A review made entirely of decade-old papers signals you missed the current conversation.

The recurring failures are predictable, which means they are avoidable:

Mistake	Why it hurts	Fix
Summary without synthesis	Reads as a reading log, not an argument	Start paragraphs with ideas, not authors
No link to research questions	Review feels disconnected from the study	End with an explicit, named gap
Author-by-author organization	Prevents comparison	Reorganize by theme using the matrix
Outdated sources only	Signals an incomplete search	Add recent work; keep only truly seminal old sources
Over-quoting	Buries your voice under theirs	Paraphrase; reserve quotes for exact wording that matters

Frequently asked questions about writing a literature review

How long should a literature review be?

For a bachelor’s thesis, roughly 8–15 pages or about 20–30% of the whole document; for a master’s thesis, 15–25 pages. A standalone systematic review can be longer. Always defer to your department’s specific guidance over any general number.

How many sources do I need?

There is no fixed rule, but a rough guide is 15–30 quality sources for a bachelor’s review and 40–60 for a master’s. Quality beats quantity: ten well-synthesized, peer-reviewed sources outperform forty listed and unexamined ones. Prioritize peer-reviewed journal articles with a DOI.

Can I use AI to write my literature review?

You can use AI to summarize sources you have already found, suggest a thematic structure and polish wording — that is generally accepted at most universities when disclosed. You must not let it invent citations or submit its output unedited. AI is notorious for fabricating plausible-looking references, so verify every source against the original. See how the leading AI academic writing tools compare on exactly this reliability point.

What is the difference between a literature review and a theoretical framework?

A literature review maps what the field has found; a theoretical framework selects the specific theory or model your study will apply. The review is broader and comes first; the framework is the lens you choose from it. In many theses the framework grows directly out of the review’s final section.

Where do I put the literature review in my thesis?

Usually as the second chapter, after the introduction and before the methodology. In shorter papers it can be folded into the introduction. Its position is deliberate: it must justify the study before you describe how you carried it out.

Key takeaways

Knowing how to write a literature review comes down to one shift in habit: stop writing about sources one at a time and start writing about ideas that several sources illuminate together. Search systematically, extract into a synthesis matrix, organize by theme, write synthesis sentences that lead with the idea, and close by naming a precise gap your study will fill. Do that, and the review stops being a chore you pad out and becomes the chapter that makes the rest of your thesis inevitable. Reference standards like the APA Style guidelines keep your citations consistent while you focus on the argument itself.