Template
Turn results into an argument your reviewers will accept
Your results are written up and you now have to say what they mean without claiming more than the data will carry.
- Typical length
- Around 20–25% of a journal article, and one of the longest chapters in a thesis. It is usually the section that grows most between drafts.
- Usual tense
- Past tense for your findings and for what prior studies did; present tense for interpretation, implications, and what is now understood.
- First person
- Frequently used here even in journals that avoid it elsewhere, because interpretation is explicitly the authors’ own.
- Hedging
- Expected. “Suggests,” “indicates,” and “is consistent with” signal calibrated confidence; “proves” and “demonstrates” rarely survive review.
- New citations
- Acceptable when a comparison genuinely requires them, but a discussion full of first-time references usually means the introduction was incomplete.
- Check with your department
- Whether discussion and conclusion are separate sections, and whether a dedicated limitations subheading is expected.
THE SKELETON
Section by section
Each block gives you the heading, what it has to accomplish, and the questions that fill it.
1 · about 10% of the section
Key Findings Restated
Open by answering the research question in plain language, without statistics. A reader who skipped the results should still follow the rest of the section.
- In one sentence per research question, what did you find?
- Are you repeating numbers here that the results section already reported?
- Would your supervisor recognize this as the study’s headline?
2 · about 25% of the section
Interpretation
Explain why the results came out as they did. Offer the mechanism you believe in, and name the competing explanation you cannot rule out.
- What is the most likely explanation for the main finding?
- Which rival explanation survives your data, and why?
- Does the theory you started with account for this, or does it need adjusting?
- Is there a finding you have quietly left uninterpreted?
3 · about 20% of the section
Relation to Prior Work
Place your result on the map drawn in your literature review. Agreement, extension, and contradiction each need a different kind of sentence.
- Which published finding does yours replicate, and in what new setting?
- Where do you disagree with earlier work, and what might explain the difference?
- Have you cited anyone here who never appeared in your introduction?
4 · about 20% of the section
Implications
Say what follows — for theory, for method, and where the field expects it, for practice or policy. Keep each implication traceable to a specific result.
- What should a researcher in this area do differently after reading this?
- Is there a practical or policy consequence, and who would act on it?
- For each implication, which result is it based on?
5 · about 15% of the section
Limitations
Name the constraints that genuinely bound the conclusion, and say what each one does to the claim. Honest limits strengthen a paper; ritual ones waste a paragraph.
- Which limitation would a reviewer raise first?
- For each limitation, what specifically can you no longer claim?
- Are you listing a weakness without saying what it costs the argument?
- Did anything unexpected in data collection constrain the result?
6
Future Directions
Point to the next study specifically enough that someone could run it. Avoid the generic call for further research.
- What is the single most valuable next study, and what would it measure?
- Which limitation could that study remove?
- Is there a dataset or method that has become available since this study began?
7
Closing Statement
One short paragraph stating the contribution at its honest strength. In many papers this doubles as the conclusion.
- What does the field know now that it did not before?
- Is the verb in your final sentence supported by your evidence?
- Have you introduced any new idea here that the paper never tested?
SENTENCE STARTERS
Phrasing you can adapt
Lift these and change the specifics. Academic phrasing is a shared convention, not something each writer has to reinvent.
This study set out to determine whether … The results indicate that …
The central finding is that …
Three results stand out from the analysis.
Overall, the data provide moderate support for …
This finding is consistent with Author (Year), who reported …
In contrast to earlier work, we found no evidence that …
Our estimate is smaller than that of Author (Year), possibly because …
These results extend Author’s (Year) account to a setting in which …
Because the design is cross-sectional, the direction of the relationship cannot be established.
The sample was self-selected, which likely overstates …
Reliance on self-report leaves open the possibility that …
These constraints mean the results should be read as … rather than as …
The evidence is consistent with, though not conclusive for, …
It seems reasonable to conclude that …
This may reflect …, but the present data cannot distinguish between …
Further work is needed before … can be recommended.
Questions about the discussion section template
How do I avoid repeating the results section?
Results answer “what,” the discussion answers “so what.” If a sentence contains a test statistic or a raw count, it probably belongs above. Refer to findings by name rather than by number.
What if my results contradict my hypothesis?
Report the contradiction directly and treat it as the interesting part. Look for explanations in the design, the measure, and the population before concluding that the theory is wrong — and say which of those you can and cannot rule out.
How many limitations should I list?
Two to four real ones, each with a consequence attached. A long list of small caveats reads as defensive and buries the limitation that actually matters.
Should the discussion and conclusion be separate?
It depends on length and venue. Short papers often close the discussion with a concluding paragraph; theses and longer articles usually give the conclusion its own section for implications and future work.
Related templates
Research Paper Outline
You are writing an empirical paper for a journal or a graduate course and you want the structure settled before you draft a paragraph.
Case Study Report
You are analyzing one organization, patient, community, or event in depth, and the report has to end somewhere other than “this was interesting.”
Lab Report
You completed a practical and have to write it up in the standard scientific format, usually against a marking rubric.
Response to Reviewers
You received a revise-and-resubmit decision and now have to answer every comment in a way that gets the paper accepted.
An outline is a start
Plan mode is the rest
Copyeditor’s Plan mode turns a structure like this into sections you produce one at a time, with citations landing in the bibliography as you go.
