The Discussion Section Tells Us Where We Are

In the Discussion section of a research report, the study's results are situated for the reader in the context of past and future research, clinical evidence, and theory. The author must look forward and backward to demonstrate to what degree, if any, the present work has moved the science forward, and where the field should go next. This section may require more thought and effort than any other part of a manuscript. If science is a long and winding road, and research is how we move forward on that road, then the Discussion section should show the reader on a roadmap (or GPS system) how far we have moved after the present study, if we have moved at all. Keeping this principle in mind, writers can craft accurate and useful Discussions. Many authors begin the Discussion by summarizing the study's results. This is rarely necessary. Although the author may have produced the results many months ago and believe that reminders are needed, the reader has just read them in the preceding section. If the results were extremely complex and occupied many pages of text, it may be helpful to clarify the most important findings, as discussed later, but in a terse manuscript, it is frustrating for readers to re-read what they have just read. To emphasize a point or explain a puzzling outcome, authors may offer findings from an aspect of the project that was not described in the methods or results sections. This irritates readers who thought they already had been given an understanding of the project as a whole and do not have enough information to evaluate the credibility of the new data. Perhaps most commonly when a study is a multi-year project, the Background section of the report may have reflected only what was known at the outset of the project. In some journals’ stylistic traditions, the Background is limited to the narrow question pursued in the study, and almost all related literature is saved for the Discussion. Either way, when the Background has not provided the state of the science on which the present work was based, to learn of it later can frustrate an invested reader. Reviewers and editors are particularly unhappy when authors claim some kind of importance for findings that fail to reach the previously identified indicator of statistical significance, are in fact only a very small contribution to the literature, or could have been predicted by someone familiar with the field. This occurs in both qualitative and quantitative reports. It conveys wishful thinking or naiveté at best, and deception at worst. Another version of this risk is claiming to have conducted the first study of its kind. This claim does not augment the importance of the contribution and cannot reliably be known, regardless of the level of expertise and seniority of the authors. On the other hand, authors may be so close to their work that they fail to see a contribution to the larger understanding in the field that is worthy of discussion. Pointing out lessons learned from a failed intervention or lack of support for a hypothesized relationship may enable others to avoid pursuing the same path. By the same token, an atypical or unexpected finding may hold the key to an important new direction. When the relevance of the present work to the larger field is not accurately portrayed, readers may doubt the authors’ depth of familiarity with the field. Over-reaching is much more common than under-reaching. Authors may see this content as required, or feel an obligation to provide some kind of useful recommendations in return for their own and others’ investment in the study. One example of over-reaching is presuming causality based on statistical association. When the recommendations are to conduct studies for which a foundation does not exist or use clinical practices that were not actually tested, reviewers and editors worry a great deal, because readers who skipped the method section or lack training to understand it may take unsound recommendations at face value. Even when work is clearly groundbreaking and important, readers should beware when the author indicates that they “must” change an approach or adopt a practice. One study is rarely sufficient evidence on which to base a change. On the other extreme, when authors offer no implications of their work for future research or practice, reviewers and editors may question the authors’ familiarity with the pressing issues of the discipline. This can occur when the authors are not immersed in the study's field, as when social scientists study nursing practice because they have access to data but are not immersed in the literature of the discipline. In an effective Discussion, the author takes a step back to show the reader the big picture of how far this study has moved knowledge forward. To gain this view, readers need a wide-angle perspective of their present location on the map of the field. Readers are grateful when authors clearly and concisely depict how the results do or do not move the science forward. Readers’ confidence grows when the Discussion provides a road map of where they are after this study, in comparison to where they were before. To provide this accurate appraisal, authors should remind readers of the project's overall goals, and the limits of the approach taken to reach them. Whether the goals were modest, such as to validate an instrument or determine the opinions of a small target group, or grand, such as conducting a meta-analysis of the effects of a widely used clinical approach, authors should be transparent in their reporting of what was done and what they realistically could hope to achieve. To further stretch the roadmap metaphor, providing a realistic picture of the present study helps readers remember whether they are traveling in an old subcompact car with many hungry children aboard or a sleek high-speed roadster without passengers. The distance we can hope to travel in one vehicle is very different than in the other. This reminder of the study vehicle can be as simple as starting with, for example, “This descriptive correlational study was adequately powered and with a sufficiently diverse sample to achieve its aims of determining statistical relationships between (conditions of interest) and (outcome of interest). The findings may provide a new perspective on (condition) and can serve as a step toward evaluating the potential for an intervention to modify (condition of interest) in hope of creating improvement in (outcome). Nonetheless, caution is needed in interpreting the results, given the nature of the sample, measures, and design.” This type of Discussion opening helps readers step back from the minute facts of the results just presented and readies them to learn what can be gained from these results in the context of the study's purpose, design, and methods. The context of a study incudes the scientific, clinical, social, and political, and sometimes epistemological worlds to which it might contribute. All are worth considering for comment. A common first paragraph of a Discussion section is a description of how the findings support or contradict past reports. There is no obligation to mention every past report with which the results are congruent and every report with which they are not, even if all were cited in the Background. Better to provide a synthesis and summary of how the results should be judged, with citations as appropriate. If the results were congruent with all past reports, this can be stated simply and briefly. If this congruence is surprising, given improvements in the vehicle in the present study (e.g., stronger design, sample, or measures), those points are helpful. If a result contradicts past reports, authors can discuss the degree to which the finding is trustworthy (not threatened greatly by statistical or other error). Then they can discuss why that may have occurred and whether it is important as a step forward along the scientific path. The scientific context of any research is complex and multifaceted. Deciding which aspects of the present study are important and helping readers understand that judgment is the most challenging aspect of framing a Discussion. To give readers the broad view they need, authors can remind them of how the present work fits into the major initiatives in the field. If the present work was an exploratory side trip off the main highway of discovery, authors can remind readers of that main direction and indicate the likelihood, if any, that the current project could offer an alternate route. If the effort led to a dead end, this is valuable to know. If the work makes a meaningful contribution to progress down the main road, authors should depict this and justify it based on the rigor of the present work. The typical study produces incremental progress. This is valuable, given the complexity of knowledge development, but readers should be reminded to zoom out from the close-up map view and remember how far there is yet to go. From the scientific context, authors can move to the clinical and social contexts and beyond. Maybe an exciting statistical discovery really has no impact at all on business as usual outside research. That is perfectly fine. On the other hand, If a study was designed to depict the extent of a problem in a population not previously studied, test a less expensive treatment that was found less effective, or try an approach that was found effective but would be prohibitively expensive to use in practice, those results should be placed in their social and clinical contexts. Similarly, the political context of results is important when they indicate, for example, that prevailing health policies are doing harm. A strong qualitative study can contribute to the epistemological context of care, when an objective clinical indicator of a health problem is convincingly supplanted by an emotional or social view of it from the experiencing subjects. Authors have the responsibility to recognize the big picture surrounding their research ventures and direct the reader to appreciate accurately the scale and significance of the contribution. Readers are accustomed to seeing a section of the Discussion on limitations of the study's scope, design, methods, and enactment—the study's vehicle, its drivers, and the road conditions they encountered. Would this result have occurred under different conditions? Would I have taken so long to get home if I were in my fast new car at midnight, and not on my old motor scooter in rush hour? The Discussion should provide readers with interpretation of the results while taking all relevant context into account. Certainly the study conditions must shape readers’ understanding of its contribution to progress, but if relegated to a section near the end, the reader may struggle to evaluate them. In my view, limitations are easier for readers to grasp when woven into the appraisal of the study in its scientific context as described above than in a separate section. Specific limitations affect interpretation of specific findings, and the author must help readers link them. The easiest way to do this is to present the degree of progress achieved, whether large, small, or nonexistent, along with the reminder about the vehicle being used to achieve it. A research journal like RINAH exists mainly to provide scientists with building blocks of evidence that can be foundations for future studies. Realistic and thoughtful recommendations for further research in the area are very important. Authors may have recommendations for design and methods or for questions or hypotheses. When the authors are the best positioned in the field to say where to go next, under-reaching in research recommendations greatly diminishes the value of the article. To avoid over-reaching or under-reaching, authors must make a calm and clear-eyed appraisal of what was and was not accomplished in the present study and whether to continue in this direction or head in another. If one's perspective has shifted over the course of the study, this insight is valuable and should be shared. After acknowledging weak measurements and an inadequate sample that made answering a research question impossible, it may not always make sense to recommend redoing the study with better measures and a larger or different sample. Perhaps the science has taken a new direction in the meantime that is more feasible or profitable. Again, taking a step back to appraise the landscape enables the author to guide the reader along the road. RINAH articles are not required to include clinical recommendations, and clinicians are not our primary audience. Given the incremental nature of much research in nursing science and the great vulnerability of patients whom nurses serve, recommendations for practice require careful deliberation. Here, the assumption that association equals causality is dangerous. The temptation is great to recommend widespread adoption of a behavior that was found statistically related to a health outcome and seems harmless. Unless experimental evidence is adequate, it is better to be safe than sorry. A lesser risk but still worth mentioning is the common recommendation that an assessment measure only tested in research samples be used in clinical settings. Clinical studies are needed to establish its predictive validity before it is put to widespread use. Nonetheless, many studies can provide food for thought for clinicians. In some cases, one can sensitize clinicians to possibilities to consider and suggest they be alert for the kind of definitive research still needed. The deliberation reflected in a Discussion section is an essential component of a fully considered piece of scientific work. Whether they were driving a hugely expensive randomized controlled trial or a tiny exploratory case study, the authors’ careful synthesis of where we as readers were before this study, where we are now, and caveats on how we got there are critically important for those who might seek to use the report to guide next steps. Margaret H. Kearney Editor University of Rochester Correspondence Margaret H. Kearney, University of Rochester, Rochester, New York. Email: [email protected]