How to Write a Literature Review: A Step-by-Step Guide for Grad Students

If you are trying to work out how to write a literature review, the first thing to know is that almost everyone gets the genre wrong the first time. The UNC Writing Center opens its handout on the topic with a common misconception: a literature review is not you reading a few books and giving a thumbs up or thumbs down. The “literature” is the published scholarship on a topic, and a review is a structured discussion of it, not a reaction to it.

This guide pulls together advice from three non-commercial sources: the Purdue Online Writing Lab (OWL), the University of North Carolina Writing Center, and “Ten Simple Rules for Writing a Literature Review,” a 2013 paper by Marco Pautasso in PLOS Computational Biology, written from his experience working on about 25 reviews as a PhD and postdoctoral researcher. Links are at the end.

What a literature review actually is

Purdue OWL defines a literature review as a document, or a section of one, that collects key sources on a topic and discusses them “in conversation with each other.” That last part, the conversation, is what separates a review from a stack of summaries.

A review can stand alone as a paper or an assignment, or it can sit inside a research article between the introduction and the methods. Either way, it usually combines two moves that UNC’s handout distinguishes carefully:

• Summary is a recap of what each source says.
• Synthesis is a reorganization of that material: grouping sources by theme, tracing how thinking in the field has changed, connecting findings that agree and naming the ones that fight.

A review that only summarizes reads like a book report. The synthesis is the actual contribution.

Step 1: Clarify the assignment and find models

Before searching for a single paper, UNC recommends getting answers to a few boring questions from your instructor or advisor: roughly how many sources are expected, what types (journal articles only, or books too), whether you should summarize, synthesize or critique, and whether the review needs subheadings or historical background. Fifteen minutes of clarification saves a week of rewriting.

Then find two or three published reviews in your field and read them as models. UNC’s practical tip: put the word “review” next to your topic in a database search, and mine the bibliographies of papers you already have. You are looking for how reviews in your discipline are organized, because conventions differ. Purdue OWL notes that humanities reviews often carry overt argumentation, while science reviews report study designs and results more neutrally.

Step 2: Narrow the topic before you drown in PDFs

There are hundreds or thousands of papers on most topics. UNC’s advice is blunt: the narrower your topic, the fewer sources you need to survey properly, and your instructor does not expect you to read everything that exists.

Pautasso’s Rule 1 adds a useful filter for whether a topic is well-chosen at all: it should be interesting to you, important to the field, and well-defined enough that you are not implicitly including thousands of publications. If your topic fails the third test, shrink it until it passes.

Step 3: Search systematically and take notes while you read

This is where reviews quietly fall apart. Pautasso’s Rule 2 is “search and re-search the literature,” with practical habits worth copying:

• Keep track of the exact search terms you used, so your search can be replicated (and described later in the review).
• Keep a list of papers you cannot access immediately, so you remember to chase them down another way.
• Use a reference manager from day one, not after you have 60 unnamed PDFs in a downloads folder.
• Define exclusion criteria early: what makes a paper irrelevant to your review? Writing this down sharpens your scope.

And take notes while you read (his Rule 3), not after. Purdue OWL suggests going one step further and building an annotated bibliography as you research: a few sentences per source on what it argues and how it matters. Much of that text transfers directly into your draft, which means you are partially writing the review during the reading phase instead of starting a blank page at the end.

Honest mention: this reading-heavy phase is exactly where sustained attention becomes the bottleneck, and it is the problem we built Refocus around. Timed focus sessions and a task broken into small steps (“read and annotate three papers”) work better than an open afternoon labeled “lit review.”

Step 4: Organize around ideas, not authors

The single most common structural mistake is organizing the review source by source: paragraph one about Smith, paragraph two about Jones. UNC is direct about this: a literature review is usually organized around ideas, not the sources themselves. As you read, look for what connects your sources. Do they agree on one solution or split into camps? Is there a debate, a trend, a gap nobody has addressed?

Once you have a theme, pick an organizational pattern. Purdue OWL lists the main options:

• Chronological: trace how the topic developed over time. Good for unfamiliar topics, but dangerous, because it tempts you into listing summaries in date order. You still need to analyze turning points and debates.
• Thematic: group sources by recurring themes or subtopics. The most common choice for standalone reviews.
• Methodological: compare what different research methods (qualitative vs. quantitative, say) have found on the same question.
• Theoretical: organize around competing theories or frameworks, common in the humanities.

Pautasso’s Rule 7 suggests drawing a concept map of your review before writing: a diagram of your themes and how they link often reveals the logical order better than an outline of sentences.

Step 5: Synthesize, don’t summarize

You will know you are summarizing instead of synthesizing when each paragraph leans on a single citation. Purdue OWL’s heuristic: when a paragraph cites multiple sources at once, you are more likely synthesizing. Their memorable framing is that synthesis is like hosting a dinner party: imagine all your sources in a room discussing your topic. What are they saying to each other? Where do they agree, interrupt, contradict?

Pautasso’s Rule 6 raises the bar further: a good review is critical, not a stamp collection. After reading it, your reader should have a rough idea of the field’s major achievements, its main areas of debate, and its outstanding research questions. That third item, the gap, is often the whole point, because your own research proposal will aim at it.

Step 6: Decide what kind of review you are writing

Not all reviews are the same genre, and Pautasso’s Rule 4 is to choose deliberately. A mini-review focuses on the last few years with tight word and citation limits; a full review covers a development in depth. A narrative review is qualitative; a systematic review follows a predefined protocol for gathering evidence to reduce bias (and becomes a meta-analysis when it pools quantitative results statistically). For a course assignment or a thesis chapter, you are almost certainly writing a narrative review, but knowing the vocabulary keeps you from promising a systematic review by accident.

Step 7: Draft, get feedback, stay current

Pautasso’s remaining advice is unglamorous and true: get feedback on a draft from peers or your advisor before submitting (Rule 8), include your own work only where it is genuinely relevant and describe it objectively (Rule 9), and stay current without ignoring older seminal studies (Rule 10). On currency, UNC adds a discipline check: in the sciences, sources even two years old can be obsolete, while in the humanities a historical sweep may be exactly what is needed. When in doubt, look at what recent reviews in your field cite.

The short version

Clarify the assignment, narrow the topic, search with notes and exclusion criteria, organize by themes rather than authors, synthesize instead of summarizing, and get a draft in front of someone before you polish it. The mechanics are not mysterious; the hard part is the hours of focused reading and writing underneath them. Protect those hours and the review mostly writes itself.

Writing a thesis or a review chapter? Refocus helps you do the deep reading and drafting in timed focus sessions, one small step at a time. Try it free at refocus.co.in.

Cover photo: Unsplash.

Sources

• Purdue Online Writing Lab (OWL), “Writing a Literature Review”: https://owl.purdue.edu/owl/research_and_citation/conducting_research/writing_a_literature_review.html
• UNC Writing Center, “Literature Reviews”: https://writingcenter.unc.edu/tips-and-tools/literature-reviews/
• Pautasso, M. (2013), “Ten Simple Rules for Writing a Literature Review,” PLOS Computational Biology: https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1003149