What are we talking about?
Here, we are talking solely about the literature review which you are expected to produce early in your PhD (or MSc) and which will comprise the first scientific writing you will produce for your thesis, not about literature reviews for publication (this is another topic). There are actually quite a few online resources on this topic, but this does tend to also be something with very field-specific expectations and conventions. This document is mainly intended as a guide for what I expect from my PhD students as a supervisor, along with some information about what is important to include and why.
Defining the scope of the problem
The hardest part about starting a literature review on a brand-new topic is working out what the scope should be. It’s very easy to get side-tracked and bogged down in huge, deep pools of literature that are not actually directly related to your thesis topic, but seem like they should be at the start. Hence, the first step towards writing a literature review is to produce an outline. This outline should have section headings with a few dot points underneath to indicate what the scope and structure of the review should look like. This is where you also need to think deeply about your thesis topic. What are the major keywords? What hypotheses do you plan to test? Do you have a reference proposal e.g. from the grant you were employed on, or did you already write one to get your current scholarship? What does this involve? Maybe we need to discuss further to solidify the original ideas, in which case I am here for this. I can also provide a draft outline if you get stuck.
Once you have an idea of what you want to do, think about what someone with no previous experience of the topic would need to know in order to understand 1) what is already known about your topic, 2) what is not yet known (the research gap), 3) why your topic is important or why we should care about it (why are we doing this research) and 4) what methods you plan to use and why, with justification from the literature. This is also roughly what needs to go into a research paper introduction, but for a thesis literature review you need to cover the research in far more detail. Your literature review should be mostly aimed at the level of a researcher in the biological or agricultural sciences who knows nothing about your topic, whereas for a research paper this can usually be targeted more to an expert audience. The reason for this is that the target audience of your thesis will include an examiner who is not an expert on your thesis topic, but who will nonetheless need to read and understand it. As well, you need to convince your thesis examiners that you have met the primary goals for the PhD: you have become an expert on this specific topic, and have made a novel contribution to the scientific literature. Of course you are at the start of the PhD not yet at this point, but by spending time reading and reviewing the literature at the start of your project you can make sure you know that what you plan to do is novel, and that you have enough understanding of your project that you can make new hypotheses if you have unexpected results, troubleshoot problems that may come up, and know why you are doing the experiments that you are doing with these particular methods. Having a good understanding of the research literature is absolutely essential to conducting and publishing research.
Outlining the literature review
So, the outline. Start broad, and then get more specific as you go, like an upside-down triangle. At the end of the literature review should be your research proposal, and after reading the literature review it should be absolutely clear how your planned experiments address an important gap (unanswered question/s) in the research literature.
First, consider if you need to introduce broad topics (such as polyploidy and interspecific hybridization), and/or if you need to introduce the significance of the crop system you are using. The first sections should mostly deal with the broad question of ‘why should we care in general about research area x’, where this is usually a very large body of literature (there should be conferences just on this research area, e.g. rapeseed, polyploidy or genome evolution). For these sections the point is not to read every research article: there will be thousands at least. Instead, try and find the most popular (highly cited) reviews on this topic, and mostly use these for references in these first sections. I probably know which ones these are and can help you find them.
As we go further into the literature review, we narrow in. Did the first sections introduce the agronomic importance of rapeseed and / or what is polyploidy? Then let’s subsequently talk about polyploidy in rapeseed, the genomic effects of polyploidisation, or resynthesising rapeseed etc., as appropriate to the thesis topic. The methods can also be covered now if necessary. I have mixed feelings about methods sections: these are not really necessary for established methods, but can still be important for novel or developing methods, or when we have a wide choice of methods which could be used to test out experimental hypotheses. If in doubt ask me, but for extremely common and well-established methods like GWAS there is no need to justify why you are using this method. Better if possible to integrate discussions of methodology into the final sections of the review (research proposal).
In the final sections of the literature review, you want to find every single study that relates very closely to your main research question/s. For each of these studies, describe what the researchers found (including with what method or study system if appropriate), and what we still don’t know. The idea is to make it very clear what we still don’t know based on previous experiments, and thereby to define the unknown area that your research aims to address.
Length?
How long should the literature review sections be? As long as they need to be! I know that’s extremely unhelpful, but some topics are going to require more background explanation than others, or will have more or less relevant literature. Try and keep the first, most general literature review sections relatively short, i.e. half a page to a page or two maximum (although you may need more for difficult concepts), but for the last sections make sure you cover all the really relevant literature in detail, whether this is only half a page or many pages worth. I would probably expect somewhere in the range of 5000 – 10 000 words, but quality is better than quantity, and you will probably know if you do the reading how much you need to cover. Or send me partial drafts for feedback and I can let you know if I think it’s too much or too little, or if anything is missing or unnecessary to include.
The research proposal
The last part, hopefully leading directly on from the literature review, will be your research proposal. This should be well-introduced by your literature review, but should also contain the following elements. Firstly, provide a brief (half to one-page max) background condensing the major points of your literature review to a logical, step-by-step argument for why your proposed research is interesting and important and tells us something we don’t already know. Secondly, clearly state your research aims and hypotheses. Thirdly, provide an experimental plan which details the methods, experimental set up or research system you will use to test your hypotheses. Fourthly, outline your expected results (briefly) and reiterate why this is of interest. Most PhDs will involve multiple projects or planned experiments: in this case, follow this same format for each experiment. There is no rule about this but 1-2 pages per experiment should cover all the information you need. The last element of the research proposal is a timeline. For this, include all experiments, and plan using a Gantt chart (look this up if you haven’t seen one before) or similar, with time broken into 1-3-month blocks. I generally use 3-month blocks (quarterly), but it’s up to you. Estimate how long you think everything will take, then multiply your estimates by 2.5. I’d you haven’t done a particular experiment, protocol or analysis before, ask someone who has how long they would expect this to take (you can also ask me for an estimate).
Timing and completing the literature review and research proposal document
That’s it! Please try and complete the literature review and research proposal by the six-month mark. If you need to start experiments immediately due to experimental constraints (seasonality etc.) I might extend this deadline, but it’s most beneficial to complete as early as possible. If you do nothing else for the first six months but write this document then I will be more than happy, and you will have already made excellent progress towards your PhD thesis. If I don’t receive this document, I will start to worry about your ability to read papers, produce scientific writing, and plan experiments, as well as if you are capable of becoming an expert on your thesis topic. In Australia, failure to submit the literature review and thesis proposal to the graduate research school (after six or twelve months, depending on the university) is grounds for failing the PhD student out of the PhD program. This is not a strict requirement here in Germany, but know that I consider this highly important! And also, please know that I am here to help with every step of the process, from providing draft outlines, recommending papers and providing feedback on your writing and proposal 🙂
PHASE Lab 