Ending well: Re-Opening the OCAR Hourglass
Nobody knows how to write good conclusions. So for the second post on conclusions in my recap of Joshua Schimel’s Writing Science (part 1, part 2), I will continue unpacking how to write a strong conclusion.
Conclusions are important but it takes a while to get them right
When pressed for time, people read the introduction and the conclusion. So far, so good. But I’ve always struggled with conclusions. Somehow I never quite got why you need them. I always felt like I had already said everything before and that’s why the conclusion felt pointless and redundant. However, I now know that I simply did not understand what the conclusion was for. It makes sense when seen as part of the academic article hourglass that Schimel discusses.
Even in grant proposals, where space is at a premium, conclusions are vital for reiterating the problem and the proposed solution. This not only reinforces your argument but also reminds reviewers of the proposal’s key points, as reviewers may not have made up their minds yet. Now it the time to convince them. Only because it’s obvious to you what your paper or proposal was about and what the “obvious” conclusions must be, that’s not necessarily the case for a first-time reader less familiar with your topic. You need a strong conclusion to recap all that and make the paper come to its resolution.
Bad conclusions
A strong conclusion or resolution fulfills the main goals of identifying the results and clarifying their implications, i.e. synthesizing what new knowledge was gained from your research.
There are three types of bad conclusions, which fail to generate this desired result:
- Weak Conclusions: These fail to adequately contextualize the results, such as merely presenting numerical values or stating the importance of findings without demonstrating their significance. They forget to “Show, don’t tell”, thus failing to demonstrate the significance of the research presented. This type of conclusion claims importance without substantiating why, leading to a mismatch between claims and evidence, thereby both diminishing the paper’s impact and making it untrustworthy. In this case, it’s easy to overreach and underdeliver because you say it’s very important and then you don’t deliver the why, which undermines your conclusion.
- Distracting Conclusions or Surprise Endings: Introduce extraneous information or new findings that were not previously discussed. This undermines the coherence of the argument. All this, if it is relevant at all, should be integrated earlier and explored throughout the paper to strengthen the overall narrative and argument. New information presented only at the conclusion distracts from the established findings and may confuse the reader. We have come across this as “the surprise ending” in another blogpost. Sudden plot twists are not allowed in academic writing!
- Actively Undermining Conclusions: These overly focus on what still needs to be done or limitations of the current study. Overemphasis on future research needs or unanswered questions detract from your study’s achievements. While it’s constructive and common in many disciplines to suggest future directions, doing so should not overshadow the present work’s contributions. Framing the study as a crucial step that enables new questions or insights ensures that the conclusion reinforces, rather than diminishes, the paper’s value. Often, it can be useful and necessary to end with new questions or new directions, but you can do that in a way that doesn’t undermine your results. You just need to frame your study as an essential pivotal prerequisite with which now we can finally ask those new central questions or that your study has given us insights into something that simply wasn’t there before (so we can – and have to – ask new follow-up questions) and reiterate the knowledge gained. Synthesizing this new knowledge is a way to pose a new open question at the end without weakening your own paper.
Effective conclusions strike a balance by affirming the study’s contributions, framing results within the broader context, and suggesting future research directions in a way that highlights the study’s role, positioning it as a pivotal contribution that paves the way for further inquiry, thereby strengthening its overall impact.
What not to put in the concluding position
Because the last words are such a power position, people are going to accept whatever you put there as your conclusion. If you put something there that is not actually your conclusion, people are going to be confused about what your conclusion really is.
For instance, don’t put the limitations in the stress position at the end of your paper. This makes it look like your conclusion is that you contribute any meaningful insight. You may need to put some limitations in your conclusion but do it early in a “But, yes” approach rather than finishing with a “yes, but” that puts the negatives as your last words.
Maybe structure it something like this: “While we were able to solve problem X, questions so and so still remain open. Or because we have shown this and that, this opens the perspective for something else.” But again, be careful not to end with limitations (or phrase the future directions to make it seem like a success). Schimel probably explains this better than I do, so buy the book. Anyway…
Don’t put any thoughts into the conclusion that don’t belong there ever! Any weak or extraneous thoughts that accidentally find themselves there will be interpreted as a conclusion and thus, it will seem like you don’t have an argument, you don’t have a conclusion, it’s not a paper worth publishing or a proposal worth funding. If you have any other thoughts, put them somewhere else, but don’t put them in the conclusion. Make sure that the punchline is very strong and focused on what you actually want people to take away.
Explain what the data means
Unconvincing and obvious conclusions, which overreach and underdeliver, only synopsise results but fail to synthesise knowledge and frame the conclusions, making it sound like they didn’t properly understand their own data or the research context. This is another common mistake when trying to re-open the hourglass.
Explain: What does the data mean? Frame your unique data point as a member of a larger group to contextualise. Explaining what data means involves interpreting individual results and placing them within the broader context of existing research. To frame a “unique data point” (whatever that may be for you) as part of a larger group, you need to draw connections between your findings and the wider body of knowledge in your field. This not only situates your work within the relevant discourse but also highlights its significance and potential implications. Step-by-step you do the following:
- Identify the Significance: Explain how your new data point contributes to, challenges, or refines existing knowledge. This might involve comparing your results with those of previous studies, highlighting similarities or discrepancies.
- Contextualize Within Broader Trends: Place your data point within the context of broader trends or patterns in the field. This helps readers understand how your findings fit into the larger picture and why they matter. It can also provide insights into potential mechanisms or underlying processes driving the observed results.
- Discuss Implications: Explore the implications of your data point for theory, practice, or further research. Consider how it advances understanding in the field, suggests new areas for investigation, or has practical applications.
- Address Limitations and Future Directions: Acknowledge any limitations of your data point and suggest how future research could build upon your findings. But not at the end like this listing would suggest. The order is off. Once you’ve gone through the points, find a way to reformulate that puts the limitations first. Then, end with the insight.
How to re-open the hourglass
One example Schimel presents is a paper that studied migratory stress in a bird species. Obviously, that insight may translate to other bird species that migrate. In a typical formulaic way, the authors (trying to write a stong conclusion) said something that came across like “this changes all of bird studies” by which they intended to reopen the hourglass. That’s a good intention but it wasn’t well executed. Experts (and probably even non-experts) can see that your insights are probably relevant for more than just your specific bird species but they can also see that this insight doesn’t revolutionize bird studies. In this case, Schimel suggests reopening the hourglass by framing the species as a member of the bigger group “birds that migrate”.
That way, they show that their results are relevant for more than just the species studied but they also don’t alienate their readers by overreaching. It’s better to state this bigger reach implicitly by saying something like “bird species that migrate, such as” (and then input in the species that you actually studied), for example. It’s kind of the same structure as when we added definitions or short explanations to help unfamiliar readers understand potentially hard-to-understand terms without offending the readers who know them. This way, you reach but it’s likely not over-reaching. But you give readers enough of a hint so they’re able to connect the dots themselves.
Don’t include patently obvious truths but also, when widening to a group, only do this if you have the references to back it up or make it seem plausible in the literature review, otherwise it maybe an unsubstantiated stretch. You should generally err on the side of over-reaching (underreaching excludes potential readers, maybe too early in the process and you can never get them back) but your work will be best received if you manage to strike a good balance and don’t over-reach too much. Think about the disgrunted reviewers!
Honestly, I’m really bad at this part but since I also study mostly one historical individual, this is very relevant. I have received the feedback that my style is too German, I’m not making any bold claims or explaining why my research matters. I often found it weird to claim how my minute insights change how we have to view the history of alchemy. When I was first asked to do it, it seemed outright bizarre to me. Now I understand that this was exactly what they wanted me to do: re-open the hourglass to show how my mini-study provides new insight for a broader group of historians of science beyond the few others who study the same historical individual as me. It makes sense. Even if all the Maier researchers would go on to cite my work like crazy, there’s maybe like 10-20 of us tops. Without reopening the hourglass, I’m just not contributing anything that will get cited by many people. Especially if I phrase it in nebulous ways that only those can understand what my research means who are already “in the know”. But I can ask what my research means if I think of Maier as an individual who part of a larger group to whom my new Maier-based insights may apply. For instance, Maier is an early modern chymist who was also an entrepreneur, so my research applies to the early modern economy somehow, even if it’s just a small contribution. I think this will still take some time to sink in. But I’m really wondering why nobody had ever explained this to me before. I feel my American colleagues just naturally seem to know this stuff (they probably learned it in their PhD programmes). But just knowing it yourself doesn’t mean you can teach it. And that’s where Schimel comes in. Thank God! Finally I can understand what the mean American reviewers mean in their reviews (that were, honestly, often times completely cryptical to me in the past!). But they were speaking Schimel (or a variation). Now I do, too.
In the introduction, you may be overselling and then under-delivering in the conclusion if you’re not explaining what the impact of your study is. In the conclusion, you need to use the results or findings of your main part and distill them into new insight or knowledge that advances the field.
Famous last words
The conclusion (on the article level) and the last word (on the sentence level) should be your most memorable. And coincidentally, constitute the take-home message. They should be both well-written and also contain important insights. For example, in history, how the whole period (or a certain class of people or whatever) has to be seen a little bit differently thanks to the evidence that you discussed in the main part of the article. A friend told me to ask not only “So what?” but also “What does this buy me at the history store?”
To hammer your new distilled knowledge home, one would usually go back through the OCAR structure (opening, challenge, action, resolution) from the beginning by stating what actions were taken, go back to the question posed in the challenge section, and then say what your conclusion was that answers this question or in some way advances our understanding. Even if you “didn’t have results” (or not the ones you were hoping for), you probably have some sort of outcome. You need to state this clearly and not bury the lead.
That’s it for today. And thanks for all the fish!
S
OpenEdition suggests that you cite this post as follows:
Sarah Lang (May 31, 2024). Ending well: Re-Opening the OCAR Hourglass. Epigrammetry. Retrieved February 16, 2025 from https://doi.org/10.58079/11r26
This was an interesting read, but the author’s last name is “Schimel”, not “Schimmel”.
Thanks for spotting that! That’s what you get from listening to Audiobooks 😉 I’ll try to fix it whereever it is mentioned…