How to Write a Literature Review: Practical Steps

If you have open a pile of PDFs and you have no idea what to do, you are not alone. But a literature review is just a synthesis and should be persuasive. In simple terms, a literature review should show what is known, what is not known, and where agreement and disagreement can be found. Your literature review should not be a summary of the studies and focus on the similarities, differences and gaps between studies. It should show the reader a direction for your study. (for a concise definition and expectations, see the Purdue OWL guide to "Writing a Literature Review")

Good starting points are making the scope of the review clear. Your literature review should focus on a question and limits to the paper review including dates, key studies, population, method and setting. Doing this a priori will help you avoid making decisions later, and be able to explain and defend your choices. Plan a search strategy and keywords and synonyms. Keep a record of each database or source searched, search terms, and decisions made to include or exclude. This is not a systematic review, but a little transparency is good for credibility, and will help you keep a consistent literature review structure. For practical advice on scope, organization, and tone across disciplines, see the UNC Writing Center "Literature Reviews" handout.

Next, you can then search and pick the documents and do a quick read of titles and abstracts. Then downselect to what you will read in depth. Make comparative notes as you read, not summaries. Compare the problem, theories, methods, samples, measures and conclusions. Most importantly, compare the studies. Do the findings agree? Do they disagree? Do they extend previous findings? This comparative focus is what will make a literature review a thematic literature review.

Draft the literature review using one approach and don’t switch models. Thematic literature reviews are the most flexible and can be used for any topic. The choice is between an issue or concept and could include “measurement”, “intervention”, “equity and access” and so on. In these sub-headings, you can compare several studies, starting with a claim and backing it up. For example, you might write “Recent results have converged on X, however, results differ for Y because of Z”. If you choose a methodological structure, you should evaluate the design and methodology of the studies and consider the strengths and weaknesses of different approaches. A chronological literature review works well when there are major turning points in the literature. Do not just list the events, make it clear why the literature has changed at each point. If you choose a structure, make it familiar to the reader. For example, a brief introduction that defines the scope and gives an overview. Then a body of thematic sections that build a synthesis. A conclusion that provides an overview of where the literature stands and also where it should go.

Your introduction should do three things. It should provide a brief definition of the topic and clarify what specific aspect of it you will cover. It must also make clear the inclusion boundaries and criteria you are using. It should also give the reader your one sentence take or through line of the literature and the way in which it has been opened.

For the body of the literature review, a good rule of thumb is to use three to six sources per paragraph. Compare the research questions, measures, designs and results. Use signposting. For example, use “however”, “by contrast”, “taken together” and “a gap remains” to help the reader follow the logic. Critically appraise the studies. For example, point out a few weaknesses such as small samples, confounded designs, self-report data on sensitive topics or a lack of external validity.

When writing a literature review it is useful to try to group the studies under different camps. If there are several theories, they should be clearly identified and distinguished. It might be useful to group the literature by qualitative and quantitative studies. If you have a clear through line, the literature review should have similar points in each paragraph. If you have a thesis or dissertation then it should be the basis for a research question or hypothesis.

Two common mistakes to avoid are, “annotated bibliography syndrome” which describes a one study per paragraph approach that does not compare studies and is not a synthesis. The cure is to group the studies and start each paragraph with a claim. And the other is saying that a literature review is the same as a systematic review. The difference is that a systematic review uses a literature search and protocol to provide a highly methodological approach to answering a narrow question. A literature review may be more thematic and interpretive. And to the extent that you are providing an outline of how to structure a literature review for your paper, you can see what a good outline of a literature review looks like. Two long tail aspects that can be added to your outline are how to start a literature review, how many sources in a literature review and an outline of a literature review. These topics are good for SEO but also provide good information.

Do not read papers for the sake of adding them to a bibliography. Read them with a thought towards making a synthesis. It is unethical to fake or pad your references, and attribution is mandatory. You can read them yourself. If you want to use a reference manager, feel free to do that. But reference managers are not a substitute for good critical reading. You can use tools like an Excel matrix to structure your synthesis. You can use a style polisher to improve the readability of your writing. But the tool must not alter the author’s voice, change the arguments or misrepresent the citations. If you want a product polisher, you could use GPTHumanizer.