How to write a discussion section for a research paper comes down to answering one question you haven't been asked yet: so what? Your results section told the reader what happened. The discussion section tells them why it matters, and that's a different kind of writing than most students have practiced before. This guide covers what belongs in your discussion section, how it's different from results and conclusion, and how to write about your study's limitations without undercutting your own work.
What Goes in the Discussion Section
The discussion section interprets your results. It answers your original research question directly, explains what your findings mean, compares them to existing research, and names the limits of what your study can actually claim. According to Purdue University's Online Writing Lab, a discussion section moves from specific to general: start by interpreting what you found, work through each hypothesis or research question in depth, compare your findings to existing literature, and close with the questions your study leaves open.
Here's how the discussion section relates to the sections around it:
| Section | What It Answers | What Doesn't Belong |
|---|---|---|
| Results | What did you find? | Interpretation, comparison to other studies |
| Discussion | What does it mean? | New data you haven't already reported in results |
| Conclusion | What's the one-sentence takeaway? | A second round of detailed interpretation |
A discussion section research paper writers get right treats this as the section where the actual argument of the paper lives. Results proves you did the work. Discussion proves you understood it.
Discussion vs. Results: Where the Line Actually Is
If you already wrote your results section, you know the rule that governs it: report what happened, save interpretation for later. Discussion is where "later" arrives. "This suggests the intervention improved retention" only belongs here, after you've already stated the number it's built on somewhere earlier in the paper.
The NIH National Library of Medicine's guidance on biomedical scientific writing is direct about the boundary in the other direction too: don't repeat results verbatim, and don't introduce a finding in the discussion that never appeared in your results section. If a number shows up here for the first time, it's in the wrong place.
A useful test: read a sentence from your discussion section on its own. If it's a fact you could point to a table or figure for, it's a results sentence that wandered into the wrong section. If it explains why that fact matters, what caused it, or what it connects to, it belongs here.
How to Interpret Your Findings Without Overclaiming
The same NIH guidance describes discussion sections as following an inverted-cone structure: narrow at the top, wide at the bottom. Your opening paragraph directly answers your research question and states your most important finding. The middle works through each hypothesis or sub-question, organized by scientific logic, most important connection first. The closing paragraph widens out to what your study means for the field and what should happen next.
Tying Every Claim Back to Your Research Question
The most common overclaim in a first research paper is treating "the data changed" as proof of "why it changed." A correlation between two variables doesn't establish which one caused the other, and a single study, especially one run over a semester with a small sample, rarely proves anything as sweeping as "this method works." Aspire's position on student research holds here directly: the value of a project comes from doing your own analysis honestly, which means stating what your specific data supports and no more. Say "this result is consistent with X" instead of "this proves X," and say "in this sample" instead of implying your finding applies everywhere.
Discussion vs. Conclusion: Are They the Same Section?
Not exactly, though in a short student paper they're often merged into one. Purdue OWL's structure treats the discussion's closing paragraph as the place where you widen out to implications and future directions. When a paper has a separate conclusion section, that section is shorter and narrower: a paragraph or two restating your main finding and its significance, without repeating the point-by-point interpretation you already gave in discussion.
If your assignment or competition doesn't specify separate sections, folding conclusion into the end of your discussion is standard and expected. If it does require both, the test is length and depth: discussion does the interpreting, conclusion does the closing.
Addressing Limitations Honestly
Every study has limitations, and naming them is not the same as weakening your paper. The NIH guidance is explicit: "no study is perfect and every study has its limitations," and the discussion section is where you show you understand which ones matter for your specific findings. A results section padded with hedges reads as unsure of itself; a discussion section that names its limitations plainly reads as credible.
Common Limitations in a First Research Project
A small sample size limits how confidently you can generalize your finding beyond the people or cases you studied. A short study window means you can't speak to long-term effects. A self-selected group of participants, students who volunteered, for instance, may not represent the wider population you're interested in. Working from an existing dataset instead of collecting your own means you're bound by whatever that dataset already excluded. For each one that applies to your project, state it, explain what you did to reduce its effect if anything, and say plainly how it might have shaped your results.
How to Write Your Discussion Section
- Reread your original research question and your results section. You're about to answer the first using only what you established in the second.
- Write one paragraph that directly answers your research question. State your most important finding first, before any supporting detail.
- Work through each hypothesis or sub-question in turn. Explain what the data suggests about it, and connect it back to why you asked the question in the first place.
- Compare your findings to existing research, naming specific studies where your results agree or disagree with what's already known.
- List your study's real limitations and, for each one, one sentence on how much it should make a reader discount your findings.
- Suggest one or two specific directions for future research, tied to a gap your own study couldn't close.
- Read it back and cut anything that only repeats a number from your results section. If a sentence doesn't interpret, compare, or contextualize, it doesn't belong here.
Frequently Asked Questions
What is the discussion section of a research paper?
The section where you interpret what your results mean, connect them to your original research question, compare them to existing research, and state the limitations of your study. It comes after results and answers the question "so what?"
What's the difference between the discussion and results sections?
Results reports what you found. Discussion explains what it means. If a sentence isn't already backed by a number or observation you reported earlier, it belongs in discussion, not results.
What's the difference between the discussion and conclusion sections?
In a short student paper, they're often the same section. In a longer paper with both, the discussion interprets and contextualizes findings in detail, while the conclusion is a short closing summary of the main takeaway, usually a paragraph or less.
How do I write about limitations without weakening my paper?
Name it directly instead of burying it in a hedge, then say what it likely did to your results. A study that owns its limitations reads as more credible than one that pretends to have none.
How long should a discussion section be?
Long enough to address every hypothesis or research question you set out to answer, compare your findings to relevant existing research, and cover limitations and future directions. For a first research paper, that's usually several paragraphs, not several pages.
Do I need to compare my results to other studies?
Yes, if comparable research exists. Say specifically where your findings agree or disagree with it. If nothing directly comparable exists, say that instead of skipping the comparison silently.
What order should the discussion section follow?
Start narrow and end broad: open by directly answering your research question, then interpret your findings in detail, compare them to existing research, address limitations, and close with what a future study should look at next.
Can I introduce new data in the discussion section?
No. Every number or finding you discuss should already appear in your results section. The discussion interprets existing data, it doesn't introduce new data.
The Bottom Line
A discussion section does the interpretive work your results section deliberately avoided: it tells the reader what your findings mean, how they connect to what's already known, and where your study's limits are. Getting it right means resisting two opposite failures, overclaiming what a small first study can prove, and hiding behind hedges instead of stating your limitations directly. Neither serves the reader, and neither is what a mentor or judge is actually looking for.
If you want a mentor checking your discussion section line by line, the Aspire Research Fellowship pairs you with a PhD mentor and a recent ISEF champion over 12 weeks, open to grades 8 through 11, with more than 20,000 students through the program so far.
Apply to the Aspire Research Fellowship →
For the section that comes right before this one, see our guide to writing the results section of a research paper, or start from the beginning with our guide to what an abstract in a research paper actually is.