Call me old and cranky if you like, but I have read a lot of bad literature reviews in my time. I know I am in for a challenging read when the chapter or section is actually entitled ‘literature review’! I was asked to provide guidance to a group of student recently, so after a quick Google which yielded loads of information but also a lot of conflicting advice I decided to reach for my keyboard.
A literature review is a description, explanation and above all else a synthesis of existing literature on a topic. I place emphasis on the synthesis part; this is where the marks are to be found if you are a student. So what does synthesis actually mean? Well a synthesis is a combination of components that form a connected whole. A pile of bricks is not a synthesis, but a house is. Equally a chronological description of the literature is nothing more than an annotated reference list. By contrast a useful literature review is a document which: synthesises common themes; synthesises literature either side of a debate; or scopes the unresolved issues or research gaps.
It is generally agreed that there are two type of context in which you would need to review the literature. You may be set an assignment in which you are asked to review the literature on a subject. Essentially this is an essay (usually with sub-headings) which demonstrates that you have found, read and understood the scientific writings on a subject. The emphasis here is on peer-reviewed journal articles and edited chapters, not web-sources or text books, although this can vary between disciplines. In a student context is just another assignment to be conducted within the rules set out in the assignment brief or to be clarified by your tutor. In a professional context there are scientific review papers which aim to promote a discipline and define future research agenda within it, or simply to bring together a dispersed and obscure literature. There is an important sub-set here which is the systematic review. This is common in medical disciplines. For example there may have been several different medical trials of a particular treatment over a number of years. A systematic review finds all this data, reviews its quality, provides analysis of this data and draws out a conclusion about future practice from that data. They are designed to provide a complete, exhaustive summary of current evidence relevant to a research question. The emphasis is on using and synthesising published data on a subject. We are going to set them aside here because they have their own rules and methodologies.
Apart from a student assignment or a dedicated review paper the other context in which one must review the literature is at the start of a dissertation, thesis or academic paper. This is a key part of the introduction and does not need to be sign-posted with the words ‘literature review’. In my view it is crass to do so. In a PhD thesis where the scope is much greater it might form a standalone ‘state of the art’ chapter, but in almost every other context it is simply part of the introduction. The aim here is to demonstrate:
- The rationale and context for your research question demonstrating that others think it is important and timely to study as well. The work of offers is used to support your claim that it is interesting, important or timely.
- That you are not simply replicating published work but contributing to a knowledge gap or refining a question that you understand. That by contributing new data, or taking a different approach, to an unresolved issue you are advancing knowledge or practice.
- That you know what you are talking about! That you are using common methods, definitions, terms and approaches. That your reader identifies with these terms and understands any variance you have adopted from the established norms.
So let’s look at a literature review in the context of it being simply an extended introduction. There are two key statements or paragraphs. The one you start with and the one you end with! You need to start with a broad focus or ‘hook’ that attracts your reader. It is not quite a newspaper headline like the Sun’s distasteful ‘Gotcha!’ headline of 1982, but its aims are similar to draw the reader in and make an initial point. Why is the study important? Why is this paper or dissertation worth reading? What will the output be and how will it advance knowledge or practice in the area of study, or at least hopes to?
The final paragraph should loop back to the ‘hook’ and define the question, scope or aims/objectives.
“In light of the work by Tilman (2015) and Bird (2018) who collectively re-define this question and showed how it could be re-evaluated by contributing new data, it is timely to ask . . . . . As a consequence this study aims to contribute new data to this debate by . . . . . . This aim will be realised through four objectives which are: . . . . . . .”
So we open the introduction with something broad that draws the reader in and then narrows the focus down to the research question in hand. Now between these two points – the start and the finish – we need to place the work in context. We need to define key terms, provide relevant background as appropriate, scope previous work in the field to help reader understand context of your approach and how your work will contribute. This is the literature review. The overall form of the text should be triangular; you start broad and progressively sharpen and focus the argument (or point) until you hit the reader with your aim. And that aim should be sharp and make the reader sit up and say: “yes that makes sense, yes that will add something, wow that is well argued and justified.” Reviewing the literature is a means to an end, not an end in itself.
So how do I go about it?
Let’s break it down into a series of steps:
Step One: Find your literature, download and/or print the papers. Any academic search engine will do, my preference is Google Scholar. Work hard to build up the literature you can find and use the reference lists in the papers you do find to ‘snowball’ other literature.
Step Two: Read critically. By this I mean read in an engaged fashion, not just being critical of what you read. Interrogate the article for information relevant to your question and focus only on what is relevant to your question. Use what you read to guide and prioritise your reading. For example if one paper keeps being cited then it is clearly important; find it and read it. Writing one page summaries of papers can really help here (Fig. 1).
Figure 1: The one page summary of papers can really help. You can keep them in an online note book or hard copy doesn’t matter but the idea is that you shuffle them in any order or set them out to find patterns.
Step Three: Looking for patterns. How do the papers you read link together? Can you group some of them as saying the same sort of thing? What are the big themes? There are lots of ways of doing this, but think of detectives working a case in a TV crime drama. They usually have a wall or a board full of images, maps and items, may be with ribbons or lines drawn between key elements. They are visualising their data looking for connections and patterns. You can do the same. Use a white board to jot down some key themes and list the papers under each. Or use the floor or a large table and create piles. Basically any way of trying to see your literature not in date order (or in the order you read it) but in terms of a new pattern. This pattern provides the structure for review and the individual paragraphs within it. You may find some gaps in your reading at this stage and need to go back to the earlier steps. In facts steps One to Three often occur in parallel.
Step Four: Drafting. Set the papers aside and turn to your keyboard. Start by jotting down they key themes and then elaborate these into sentences and paragraphs. Cite references from memory where you can, but don’t write from the papers themselves. Write freely. It is all too easy to slip into copying sentences or paragraphs consciously or unconsciously – don’t! The reader wants to see your perspective on the literature not someone else’s. Remember the triangle as you write – start broad and end up with a sharply phrased point at the end. With your draft finished review it carefully looking for gaps and inconsistencies and remember that any given paper may appear several times under different themes.
Step Five: Refine and polish. This is all about the fine detail. Are you citing the right papers in the right places? Have you spelt the author’s names correctly? Are the dates correct and in the right order? Do all the text citations correspond to a full reference in the reference list and vice versa? Do the sentences make sense? Does your argument or succession of paragraphs build one upon another and sharpen the focus broad to point? Does you aim follow clearly from the succession of paragraphs? Success is all in the detail once you have found the big themes and pulled them out.
Frequently asked questions
So what is enough? Well if it’s an assignment then look to the brief or your tutor for clarification. Typically they may use words like ‘review the main or key sources’ or ‘review the significant literature and developments.’ They are clearly asking you to be selective in what you include. Alternatively if words like ‘comprehensive’ appear you have your answer. For a dissertation you need to be pretty comprehensive is my personal view, but remember you are not describing each paper but finding the themes and citing the relevant literature to that theme.
Can I cite material cited in other papers? The simple answer is no. Only cite and discuss what you have read yourself. How do you know that another person has read and interpreted it correctly? So if I can’t get a key source? You should try to get it or leave it out; it’s that simple.
Can I cite web sites and textbooks? It depends a bit on the subject and level but basically not often. There may be a defining manual or textbook for a subject especially in an area of practice and off course this may feature heavily in your review, but where ever possible go back to the primary sources.
Should I cite old papers? Yes just because its old does not mean it is not key to a debate or seminal in the development of a subject area. It is often harder to get older material however, many of the big publishing houses have back catalogues that stop in the mid-1990’s this does not mean it not worth pursuing. Material that is not accessible electronically is not by default rubbish. If a debate has lots of recent discussion then your focus will be there, but that is not always the case.
If a whole paragraph is based on one source where does the reference go? Again there is no perfect answer but generally not at the end. Start by saying something like: “The work of Smith (1988) is key here. They studied the question closely by conducting the following experiments . . . . The key conclusion in Smith (1988) is that . . . . . The work of Smith (1988) was developed by Thompson et al. (2005) who modelled the system using . . . .” Basically make sure the reader knows that it is Smith’s work that you are talking about by reminding them regularly. If you only have a few sources, either because that’s all that was written or because you have not found more, then the same names will be used a lot and it is what it is.
When I have multiple citations against a sentence what order should they go in? Date order and if they were published in the same year alphabetically, so: “The idea that the literature view was key to the development of academia has been widely discussed (e.g., Bogs, 1967; Simons, 1988, 1989; Briggs et al., 2002; Smyth, 2002).”