What is a literature review?
A literature review is a critical evaluation
of the current state of knowledge in a chosen academic subject1
. The word ‘critical’ is important because, when done properly, this process involves comparing, contrasting, considering the strengths and limitations of each work, finding gaps and unanswered questions. By contrast, a literature report
is an annotated bibliography of works, with an (uncritical) summary of the findings of each work. This type of writing is much more complicated than something like blogging, where you just choose a niche
and you get going.
Identifying the current state of knowledge involves searching for articles about the topic we want to review and screening them for relevance. It’s important to acknowledge that you will only ever review a sample of the literature as part of your review. Crucially, how you take that sample will depend on the type and depth of the review that you want to do.Types of literature review
A systematic review
aims to answer a specific, focused research question, such as: ‘Do behavioural interventions reduce risk of stroke?’ At the highest level, a systematic review will involve a team of people carrying out an iterative process of searching, filtering and screening, to reduce the risk of bias. Multiple databases will be searched, as well as the reference lists of papers and hand-searching of relevant journals. Often has strict inclusion criteria.
A systematic review may be quantitative
, involving a meta-analysis of RCTs aggregates the results of randomised controlled trials for example. Alternatively, this type of review can be qualitative
, where common themes running through the literature are analysed and grouped in some way.
In contrast, narrative reviews
can include all literature, including non-peer-reviewed work. We’ll focus more on narrative reviews here, but the process for all of these is similar.Dos and don’ts
The ideal review should be comprehensive; objective; repeatable; documented. A literature review that meets these criteria can be published in its own right, particularly if it tells a story, such as: How did we get to the current state of knowledge? And how has our understanding [of this subject] changed over time?
At worst, a review can be cherry-picked; biased; and designed to serve the interests of the author, not the reader.
The following wording which you might find in the ‘related research’ section of a paper can indicate that the review might be biased:
“We’re the first to use this method.”Search strategies for your literature reviewTop-down
“We’re the first to solve this problem.”
“Others have tackled this problem, but our results are better than previous results.”
Start with a search based on a focused research question. Refine the search over a number of iterations. Useful tools for this type of review include: 2dSearch, Google Scholar, PubMed
, traditional academic search engines.Bottom-up
This is a more exploratory approach. Start with a few key/seminal papers as ‘seeds’ and look at the work they cite and who cites them. Useful tools for this type of review include: ConnectedPapers, scite, TheLens and SemanticScholar. Aim to combine both top-down and bottom-up approaches to find a good selection of papers.Writing the review
In your PhD thesis, the literature review will typically appear as the first chapter. If you can write it in an objective, self-contained way, it may be publishable. Let’s have a look at the structure of an effective review.Introduction and background
This is where you set the scene and take a broad view. Define your terms and provide a brief overview of the subject. What are the seminal works? Where are the gaps in knowledge? What are the problems or questions to which this review seeks answers?
Some of this information can be gleaned from earlier review papers on the subject or from your initial learnings from your ‘seed’ papers. Some of this will arise as you are carrying out the review itself, so it’s best to write this section last.
In your final paragraph, talk about your specific aims and objectives which might look something like this: ‘The aim of this review is to identify the main methodologies for solving problem X and provide a summary of the key results when applied in the setting of Y’.Methodology
Write this first as it should provide a comprehensive description of your assumptions, constraints and search strategy with the inclusion and exclusion criteria. For example, if you are doing a thematic analysis, state this here and the reasons for doing so. It's helpful to create a diagram describing your search strategy like the example below2
This is where your critical analysis of the literature is. What was the overall quality of studies that you found? What are the emerging themes and key results? Exporting brief summaries of the participants, methods, findings and limitations from a set of papers to Excel can give you a quick and easy way of assessing and comparing studies. Discussion
This is where you relate your findings back to your aims and objectives and discuss the results in context to what was previously known about the subject. You’ll also discuss limitations (very important) and future work here. You can also bring in recent papers that might have arrived since you started writing up.Conclusion
What emerged from the review? Was there a pattern or a trend in the direction of the literature? Is there a consensus or are there still disagreements? Provide one or two key findings here. What remains to be done? Is this the starting point for your own work?
With the growing number of tools and technologies available to help researchers search, screen and apply the masses of literature in their field or area of study, literature reviews no longer have to be such a daunting prospect. Combine these tools with the structured approach to writing set out above and you can produce a solid, repeatable review with fewer headaches!
- Gooch, P. and Roudsari, A., 2011. Computerization of workflows, guidelines, and care pathways: a review of implementation challenges for process-oriented health information systems. Journal of the American Medical Informatics Association, 18(6), pp.738-748.