The discussion is a core part of the write up for any piece of research, whether it is to be published as a paper, or presented as a dissertation. It is the element of the paper where you have greatest freedom to show critical awareness of your data and how it fits in with the wider body of research in the field. The discussion has a very clear purpose which is to explain the meaning of your observations. The focus should be to help your readers to understand your study and how your data has extended understanding of your field.
Most people who have to write papers, reports or dissertations will find that writing the discussion is the scary bit- the hardest part of the work. More thought will go into writing this than any other part of the paper and yet ironically it is one of the sections that most readers will never look at in any detail. Most experienced readers of papers prefer to just look at the methods and the results and draw their own conclusions. You will also find it is the section that peer reviewers will make most comment about- they will ask you to hack your well-crafted discussion about and add in statements that suit their own prejudices, or which tone down the power of your conclusions. You end up having to do this if you want the paper published and it will annoy you intensely.
If you have been reading the other sections of this How to Write series then you will have spotted that I make regular reference to brevity and clarity. The discussion is no exception. For a research paper where the overall length is going to be maybe 4000 words, then your discussion is going to be no more than 750-1000 words. Just as the introduction is not the place to show off your encyclopaedic knowledge of the literature in your field, the discussion is not a place to expound on long-winded and clever sounding theories. Keep things simple and keep things very clearly focused upon the data and upon the evidence that is available. The most common criticism that reviewers level at discussions is that they are too long and contain too much speculation that is not backed up by the data.
All discussions are different and obviously need to be tailored to the individual needs of the paper and the data that you have generated. Despite this diversity I have found that as a basic framework for planning a first draft of any discussion the following 7 points can be addressed. Once these are in place you will have a framework that you can rearrange as necessary, add in relevant linking pieces and expand/cut specific areas.
- Briefly summarise the findings of the project. Only give highlights and avoid repeating what was in the results section. One paragraph should suffice. This is just headlining the key points.
- Now you move onto the critical analysis of your study. To get this going ask yourself ‘Has anyone done work like this before?’ a) NO- then trumpet the fact that your work is novel. Why is what you’ve done important? Write about it. b) YES- Who did it? What have you done that is different? How are you taking the area forward? Write about it
- The next question to address is, ‘Do your findings agree with the literature?’
a) NO- You need to try and explain why you disagree. Is it a methodological issue? Have you studied different populations/animals/samples? Was your experimental design better, making your findings more reliable than others?
b) YES- Comment on this as it strengthens your results. You can now use similar explanations to those given by others in the field when addressing point 4.
4.You now reach a key part of the discussion where you go beyond putting your findings in context and begin to explain them. What do your results mean- how do you explain what you have observed? This is likely to be significant element of the discussion- I would suggest at least a quarter to one third of the overall length.
5. What are the limitations of your study? This is where you critique your own approach and methodologies. ALL projects have some limitations, but you don’t want to go too far as it will undermine your own study. Reviewers are bound to ask you to expand on this, so don’t go over the top until they make you.
6. Now we start to draw things to a close and you reach the one part of the discussion where you might be just a little bit speculative. What future work should arise from your project? What is the broader impact of your findings? Do you have any specific recommendations to make re practice and policy?
7. Now you can finish off and write the conclusion(s)- one punchy paragraph to sum it up. Whatever you do don’t sit on the fence when you get to this part- make a definite statement. If you were following a hypothesis-driven approach then your data either supported or disproved your hypothesis and so definite conclusions must be possible. Personally I find that papers that end by saying ‘further work is needed to address these issues’ make me very cross and are just boring.