Most early-career researchers think the hardest part of writing a paper is the methodology or results section.
But while rigour is really important for research papers, the truth is, most researchers get trained in getting rigour right during grad school. But nobody trains them to write a great argument. So, for early-career professors, more papers get rejected because of weak Discussion and Conclusion sections than any other reason. Reviewers can spot a confused researcher from the first paragraph of these sections. They know immediately whether you understand your own research contribution or if you’re just throwing words on a page like Spaghetti. The difference between acceptance and rejection often comes down to how clearly you can interpret your findings and articulate your contribution.
So, let’s discuss the 5 simple ways to write Discussion and Conclusion sections that get papers accepted at top-tier journals or at those kickass conferences that we cherish and love in my research field (HCI).
Write your discussion like a courtroom argument
Your discussion section needs to make a case for why your findings matter, and the strongest discussions follow a logical argument structure.
Start with your strongest evidence first. Present what your results actually mean in the context of your research question. Then systematically build your case by connecting your findings to existing literature. Show where your results align with previous work, where they diverge, and why those differences matter. Address the obvious counterarguments before reviewers raise them. If your results contradict established theory, explain why. If they support existing findings, explain what new insight you’re adding. Think of reviewers as skeptical jurors who need to be convinced your research matters.
The weakest discussions jump randomly between different implications without building a coherent argument. You want to build the latter.
Keep your conclusion focused on one key message
While the discussion is a full-on argumentation for your research, your conclusion should answer one question clearly: what is the single most important thing readers should remember about your research?
Everything in your conclusion should support that central message. Start with a direct answer to your research question in 1–2 sentences. Then state your core contribution in language a non-expert could understand. End with one forward-looking sentence about what this means for the field. Resist the urge to rehash everything from your discussion or introduce new ideas. Reviewers read conclusions to understand your bottom line, not to revisit your entire study. A scattered conclusion signals a researcher who doesn’t understand their own contribution.
Think of your conclusion as your elevator pitch to the journal editor or the papers chairs of the conference.
Use the inverted pyramid structure in your discussion
I love journalistic technique. So many fun things to learn about writing from a good journalist. Anyways, I digress. Journalists use the inverted pyramid to put the most newsworthy information first, and your discussion should follow the same logic. Don’t dick around. Hit ’em in the face with the biggest punch you got. Channel your inner Derrick Lewis.
Lead with your most significant finding and its implications. Follow with supporting evidence and connections to literature. End with limitations and future research directions. This structure serves two purposes: busy reviewers can grasp your main contribution even if they just skim (which they do when they first read your paper), and it forces you to prioritize your most important insights in clear language. Many researchers bury their key contributions in the middle of long discussions, making reviewers work to find the value. Don’t make anyone work for your goodies. Your first paragraph should make a reviewer think, “Wow, this is interesting” rather than “Where is this thing going?”
Save the methodological limitations and future research suggestions for the end of your Discussion.
End your conclusion with forward momentum
Many early-career researchers get this wrong and end everything with a concise sum-up of what they did. And consequently bore reviewers to death. A little more Harry Potter, a little less Bella Swan, please. The last sentence of your paper shapes how reviewers feel when they finish reading, and you want them feeling energized about your contribution. You want to open up a whole world of thoughts to them.
Avoid ending with phrases like “In summary” or “To conclude” followed by a rehash of what you already said. You are not a writing robot. Instead, end with a single sentence that points forward. Describe the most important implication for practice, theory, or future research. Make it specific enough to be actionable but broad enough to matter beyond your narrow research context. For example: “These findings suggest managers should reconsider how they structure remote teams during organizational crises” creates forward momentum. “This study shows that remote work affects team performance” just restates what everyone already knows. And, hey, you’re wasting my time if you are telling me again what I already read. That’s just rude.
Your final sentence should make readers want to cite your work. Make it tasty.
Write your discussion first, then extract your conclusion
Most researchers write their conclusion as an afterthought, which explains why so many conclusions feel disconnected from the actual research contribution. And you’re just plain boring there, Piper Chapman.
Start by writing a deep discussion section where you fully explore what your findings mean. Work through all the implications, connections to literature, and theoretical contributions. Then step back and identify the single most important insight from that entire discussion. That insight becomes the foundation of your conclusion. Extract the essential elements (i.e., your core finding, your main contribution, and your key implication) and build a tight 3-6 sentence conclusion around them. A conclusion should reflect your best thinking, not a rushed summary under deadline pressure.
Your conclusion should just feel like the natural culmination of your discussion. It should wrap things up tightly.
P.S.: Curious to explore how we can tackle your research struggles together? I've got three suggestions that could be a great fit: A seven-day email course that teaches you the basics of research methods. Or the recordings of our AI research tools webinar and PhD student fast track webinar.
Bonus Materials
Today, we have a complete framework for writing a Discussion section for a paper, including detailed guides for the 7 paragraphs it should include (at minimum), plus an LLM prompt to get your first draft done in seconds.
