How To Write a Better Discussion Section
I have heard from a number of colleagues that the discussion section is their least favorite part of the manuscript to write. I find this unfortunate - the discussion section is the most important section of your manuscript! Consider its purpose in your paper:
It highlights the importance of your findings.
It provides an opportunity to offer interpretations of your findings.
It allows you to describe how your findings are linked to the larger literature.
It gives direction to other researchers - as well as policy makers and providers, in many cases - for integrating your findings into their work.
Reviewers and readers can interpret the meaning and value of your research in many ways, so here is your opportunity to make a case on how you think they should be BEST interpreted! Also, the discussion section is often the last part of the paper that a reader reviews, so you want them to walk away from your paper feeling interested and satisfied with your work!
Unfortunately, the discussion section often opens up opportunities for criticizing your paper. In some cases, readers may feel that you are overstating your case or speculating too much about the meaning and value of your findings. Other readers may feel that you did not make a strong case about the importance of your study in the discussion section, resulting in a manuscript that ends on a flat note. There may also be disagreement about your recommendations for future research.
Writing better discussion sections is an ongoing learning goal for all of us, regardless of our stage in career. However, there are some best practices for developing a discussion section that is concise, engaging, and informative.
The discussion section is not where you copy and paste your introduction.
The introduction should be a stand-alone literature review that provides your reader with the rationale for your study. Therefore, restating a lot (or any, in some cases) of the arguments you make in the literature review would be redundant. Do more to link your findings to the larger literature, rather than discuss how the literature calls for the need for your study.
Start your discussion section with a brief description of the findings. Avoid being repetitive and/or verbose. If you did a good job in the introduction, methods, and results section, you should only need one short paragraph to summarize your findings.
Explain your findings and ground them in the existing literature.
Were your findings expected? Unexpected? Here you state why or why not. Remember that your investigation was not conducted in a vacuum - it was informed by a larger body of research. Therefore, when you explain the meaning of your findings, you should cite previous studies that offer explanations of why you found - or didn't find - what you expected. The more you can explain your findings using existing literature, the less likely the reviewer/reader can attribute your explanation to speculation. You now have valid reasons for interpreting your findings a certain way. While tempting to only highlighting prior research that supports your interpretation, there are also benefits to offering alternate explanations of your findings that may or may not be consistent with your perspective. Reviewers may commend that you went the extra mile to minimize bias.
Be careful not to over-interpret your results. For instance, there is a difference between stating that your results could mean XYZ and stating that your results do mean XYZ, Offer possible interpretations, rather than facts.
Give clear directions to other researchers, policymakers and providers.
Select the most important of your findings and provide detail about how they should be used to influence new studies. You might not have space to write about all of your findings, so be selective. When giving suggestions for future research, write in general terms, rather than proposing specific studies in detail. Otherwise, reviewers may disagree about your specific approaches or methodologies. For instance, you may identify specific topics of future research or specific research questions/hypotheses that should be explored. You might offer specific types of methodologies if there were limits to your own study that resulted in new questions about the topic. For instance, a quantitative study may have indicated that a specific association exists between two variables, but it is unclear why the association exists. You may suggest a qualitative or mixed-methods study to further explore the topic. Another example is if you had difficulty in recruiting certain sub-populations in your study. You may suggest further research on that sub-population regarding your topic and offer tips for better recruitment strategies.
If you have ideas for future studies and/or grant applications for that you plan to undertake, you may also write about how your current findings highlight the need for the future research you plan to conduct. Then, you have a citation supporting your future work once your paper is published!
Often overlooked in the discussion section is how findings are meaningful to non-research audiences, particularly policymakers and providers. If appropriate for your discipline, you should give specific attention to these stakeholders. Policymakers and providers often lament that researchers publish their findings, but do not make it clear about how the findings should be used to change things in "the real world." Researchers often lament that they write up their research findings, but they are never translated or are translated inappropriately. You can avoid both situations by clearly stating how the findings should be used to change policies or practice in the field. Doing so will also expand the audience for your research!
Avoid berating other researchers or studies.
Your research may have accomplished something that was not possible for previous researchers. Your methods yielded findings that made the interpretation of results more clear than in prior studies. You may have accessed populations that are difficult to research. Your study may have discredited prior findings. It is okay to highlight these issues in your discussion section, but do so in a professional manner. Remember that your research contributes to a community of scholars - it is not designed to be a race where there are winners and losers. You do not want to alienate yourself from this community.
Worse - consider that in blind peer review your manuscript may be given to one of the researchers that are criticized in your manuscript. It may also be given to a reviewer who is supportive of the prior research you criticize or possibly has based their own work on the studies you have berated. This elevates your chances of having your paper rejected.
Early in my research career, I published several papers on intergenerational human services that discredited common assumptions that were being made in the literature. At the time, this was a burgeoning area of human service delivery and there was significant advocacy and enthusiasm for bringing children and older people together to address health and social service goals. Unfortunately, since my findings highlighted some of the challenges of the intergenerational model, the few papers I published on this topic have rarely been cited by other scholars in the field, despite the fact that they were published in highly visible journals with impact factors. Although my findings may have been unpopular, my ability to write about such findings in a professional, respectful manner that took care in not berating the prior literature did not make publishing these papers more difficult. We don't have to always publish popular findings, but we need to be careful about how we relate unpopular findings to the larger literature in order to get them published, since they will face additional scrutiny.
Acknowledge the limitations of your study.
No study is perfect. Editors, reviewers, and readers will understand that. However, if you do not acknowledge the limitations of your own work, this will raise questions about the level of bias that influenced your study. First, caution the reader to keep the limitations of your study in mind when they interpret your results, Then, select the most significant limitations affecting your study. It may not be necessary to write all of the limitations, just the ones that most likely impacted your results.
You should balance your discussion about the limitations to your study with strengths in your methodology. For instance, "Despite efforts to translate research materials into Spanish, consulting with Spanish-speaking researchers, and the collaboration of Spanish-speaking community members for the purposes of recruitment, few Spanish-speaking Latinos enrolled in our study." Rather than try to balance each limitation, you can always list the limitations and describe how they may have influenced your findings and then end the paragraph with "Despite these limitations, there are a number of strengths to our study that increased rigor, including..."
Consulting with others about your discussion section before you submit the paper for review will help with understanding how others may interpret your findings and/or how others will interpret your own description of the meaning and importance of your findings. In the case where you make recommendations for providers and policymakers, you may also want to consult with non-academics about your findings and how they see your findings as influencing the current state of policy and practice.
The discussion section may not always make or break your potential for having a paper accepted, but a good discussion section will help you promote your own work and make a case for advancing your line of inquiry.
Need help with publishing or advice about research? Contact Nicole today and see how you can work together to advance your career!