Speaker 1: Hello, William Levesque here from the Rehabilitation Teaching and Research Unit, RTRU, at the Wellington campus for the University of Otago. This presentation is a short introduction to how to write a discussion section. I've targeted this at PhD and Master's thesis students, so I'll be talking a bit about how to write a discussion chapter for a thesis, but also it's relevant to writing your first paper for publication in a scientific peer-reviewed journal. So to begin with, here is a broad overview about what to include in your discussion section. You want to cover what you did, what you found, what it means, and these things can be summed up to refer to why your research is important. You want to talk about the strengths and limitations of your research and make recommendations for future research. If your research is to do with health science work, then you should also be talking about the implications from your research in terms of clinical practice. Specific journals do have more explicit requirements in terms of the structure of your discussion section sometimes, so do follow those of course if you're submitting a paper for publication. But in general, these are the things that need to be covered in a discussion chapter. In terms of what you did and found, you need to be declarative, direct, and succinct. You want to cover this in a paragraph or two, or if you're writing for publication, a short journal article, this might need to be only a sentence or two in your paper. You don't want to reproduce your results in entirety in your discussion section. Just concentrate on the major findings, the stuff that is most important from your work. In terms of what your research findings mean, you want to interpret your findings rather than just report them descriptively. Reference to existing theory can be useful here. Your research findings may either expand on existing theory or be used to further substantiate existing theory. You want to talk about why your research findings are important and to link your findings to prior research. Highlight the parts of your research that are new. This may be that you have explored a new research question that other people haven't explored before, or it might be that you have found something new or produced new ideas or perspectives through your research, or it might be that your research findings provide important new evidence for existing theory, strengthening existing ideas on a particular topic. So here's an example of a paper that I published last year with my colleagues on research, reporting on some research, some qualitative research, exploring cultural issues around the barriers and facilitators for Māori people accessing pulmonary rehabilitation, which is an exercise and education programme for people with chronic lung disease. I'm just going to scroll down to the results section here, and you'll see at the beginning of the results section I have a summary of our main research findings which have been linked to the prior research in this area. So our work was about barriers and facilitators of accessing pulmonary rehabilitation, so we talk about the findings that relate to, that have previously been identified in other people's work, and we highlight other findings that arose from our work with Māori individuals where these findings are more specific to that population and present new perspectives. Right at the beginning of this discussion section we make an assertive statement that, to our knowledge, this is the first study that has explored these cultural issues regarding the influence of cultural issues on uptake of pulmonary rehabilitation, so we're highlighting what is new. So what are some of the strategies for you to use in order to identify things that you should be talking about in your discussion section in more detail? One idea is to start early. While you are analysing your results, or even during the process of conducting your studies, start jotting ideas down about what sorts of things might be important arising from your work, and you can refer back to these when you're writing your research section. It's also a really good idea to have a clear picture in your head of who your audience is for your work. Write for your audience. What are the key take-home messages that you want people to take from your research, and who are you writing for? If you're writing to a specific journal, it is useful to know who the readership of that journal is, and who the editorship of that journal is, and you want to write for that for those people as well. Consider alternative interpretations of your data. Try and think what other ways you could interpret your research findings, and either use this to strengthen your own ideas or to adapt your own ideas in terms of how to interpret your research findings. Make sure you justify your ideas either way. Another strategy might be to list all the things that you didn't know when you began your research that you have subsequently discovered for yourself. It's easy, particularly for a three-year PhD process, to forget how far you've come in terms of the development of your knowledge, and so it's useful to reflect on the things you didn't know prior to undertaking a study. Develop each of these points in your list as a discrete section in your PhD or Master's thesis discussion chapter, or as discrete paragraphs in the results section of a published article. It can be also useful to treat the background section of your thesis or your published article as the setup to your discussion, and I'll just illustrate this with this paper that I've been referring to here. So if I go back to the introduction, there is reference in here to what other people have written about barriers and facilitators of access to pulmonary rehabilitation, but how little information exists on cultural factors related to the delivery of pulmonary rehabilitation. So, right at the beginning of this paper, I have foreshadowed this discussion that we would have later on in the discussion section of this paper around the importance of considering these things and the implication they might have. So this highlights what's new, again, for this research that we were undertaking. When discussing the limitations of your research, you want to remember that all research has limitations. Examiners expect PhD and Master's theses to have limitations, they expect graduate students to have faced difficulties and problems with their method, and so examiners of theses are more concerned with how problems were dealt with than whether or not they occurred. So it's important, both in writing a thesis and in presenting research for a published article, to be frank and candid about the limitations of your research, but don't be overly apologetic because limitations are inevitable. You want to discuss the implications of the limitations rather than just highlight them. You've got to talk about why the limitations are relevant in terms of interpreting the results of your work. Some limitations are just errors and others are considerations regarding the scope of the work, but both of these things are worth discussing. Discussion of the limitations also provides you with an opportunity to identify future areas of research. You want to make general statements about future areas of research, but also provide specific recommendations about what needs to happen next, what study is needed next, what methods should be employed, what is needed to do these types of studies. And again, to come back to this paper here in the discussion section, at the end we talk about how research on indigenous healthcare is sorely lacking and there's a great need for it, and make a specific recommendation here drawing on other people's work to put out a call for an urgent need for further cost-effectiveness studies to be conducted involving indigenous healthcare interventions in order to better understand the value and to further develop these models. So you can see we've made a general recommendation and some more specific recommendations in terms of the type of research that needs to be conducted. Things to avoid in the discussion section. Avoid repeating too much of your findings without interpretation. Don't just repeat verbatim. You want to take things a little bit further in your discussion section. Avoid presenting new data in your discussion. If you feel like you want to present data that you didn't present in your results section, you might want to go back and revise your results section rather than add it to your discussion section. Avoid over-inflating the importance of your findings. As a scientist, you need to be appropriately conservative in terms of how far you can develop arguments from the data that you have. And similarly, avoid over-extending your arguments. Don't interpret beyond what can be justifiably inferred from the data. One thing to look out for is long strings of inferences, if this, then this, then this, then this, when you've moved a long way away from your original data to reach some hypothetical conclusion. Avoid raising too many tangential issues. However, this can be okay in a chapter section of a PhD thesis or a master's thesis, and that's one of the luxuries of writing a thesis. You do have a bit more room to expand on some of the more fringe ideas in your work. One more consideration. It's useful to use the discussion section of your work to publish arguments that you might want to cite later. So by cite later, I mean in terms of future research grants. It's useful to have arguments presented in a paper to justify future research ideas. But also, use your discussion section to put messages out there that you want in the world. So these could be messages for clinicians or policy makers or for students. In conclusion, you want to write a conclusion and be conclusive. The conclusion might be a separate section in your thesis or in your journal article, or it might be integrated in your discussion at the end. But regardless, conclusions should be tight and to the point. They should follow clearly from your research question. Don't introduce any new citations in your conclusion statements, but make sure that there's a nice flow of logic from your research question to your methods, to your results, to your discussion, to the things you pull out in your conclusion. Be assertive in your conclusions, but be balanced. So that's it from me on writing a discussion section for your PhD thesis, master's thesis or journal article. If you want to follow my work on a more ongoing basis, you can follow me on Twitter at DrLivac.