Short answer: Mastering how to write discussion section paragraphs requires opening with principal findings, comparing results with published literature, addressing study limitations with intellectual honesty, and articulating defensible implications.
The Discussion explains what the results mean in relation to the research question and existing evidence. It should not simply repeat the Results. According to the official ICMJE Preparing a Manuscript guidelines, authors should begin with a brief summary of the main findings consider possible explanations, place findings in the context of relevant evidence, state limitations, and explain implications for future research or practice where appropriate. Conclusions should remain connected to the study’s objectives and supported by its data.
What does the Discussion do?
The Discussion moves from observation to interpretation. It answers questions such as: What do the findings mean? How do they relate to previous evidence? What are plausible explanations? What are the important limitations? What should readers take from the study?
1. Begin with the main findings
Open with a concise synthesis rather than repeating every result. Identify the most important answer to the research question and, when useful, mention whether the primary hypothesis was supported. To see how these findings were established in the preceding phase, read our tutorial on How to Write the Results Section.
2. Interpret rather than repeat
The Results say what happened; the Discussion explains its significance. For example, instead of restating every numerical result, explain how the observed pattern relates to the study question and the existing literature. For the methodological context underpinning those numbers, revisit our guide on How to Write the Methods Section.
3. Compare with previous evidence
Discuss similarities, differences, and possible reasons for disagreement with relevant studies. Differences may arise from populations, settings, exposure definitions, outcome measures, sample size, study design, measurement methods, or analytical choices. Avoid turning the Discussion into a long literature review.
4. Explain plausible mechanisms carefully
If a biological, behavioural, social, methodological, or theoretical explanation is plausible, present it as an interpretation supported by evidence. Do not turn an untested mechanism into an established fact.
5. Address limitations explicitly
Discuss limitations that could affect validity, precision, generalizability, measurement, causal interpretation, or completeness. Good limitation statements identify the issue and explain its likely consequence rather than simply listing weaknesses.
6. Distinguish internal and external relevance
Ask whether the findings are credible for the studied population and whether they can reasonably be applied to other populations or settings. Generalizability should follow from the design and sample, not from the author’s preference.
7. Discuss implications without overclaiming
Implications may concern research, practice, policy, theory, or methods depending on the field. Keep them proportional to the evidence. A single observational study generally should not be presented as proof that an intervention will produce a particular outcome. To understand how these components tie back to the paper’s opening, consult our guide on How to Write a Research Paper Introduction.
8. End with a defensible conclusion
Link the conclusion directly to the objective. ICMJE advises avoiding unqualified statements and conclusions that are not adequately supported by the data.
A practical Discussion structure
1. Main findings: What is the central answer?
2. Interpretation: What might explain the findings?
3. Comparison with literature: How does the evidence fit with previous work?
4. Strengths and limitations: What affects confidence and applicability?
5. Implications: What should researchers or practitioners take from the findings?
6. Conclusion: What can be stated confidently, and what remains uncertain?
Common Discussion-section mistakes
Repeating the Results paragraph by paragraph.
Claiming causation from correlation or association.
Introducing major new results that were never reported in Results.
Using citations to make claims that the cited studies do not actually support.
Treating one study as definitive evidence when uncertainty remains.
Listing limitations without explaining their consequences.
Making recommendations that are much stronger than the study design permits.
How to keep the Discussion balanced
Use cautious but precise verbs such as ‘suggests,’ ‘is consistent with,’ ‘was associated with,’ or ‘may reflect’ when the evidence warrants them. Use stronger causal language only when the design, analysis, and evidence justify it. The goal is not to make the Discussion sound weak; it is to make every claim proportionate to the evidence. For structural standardizations across narrative segments, you can check the guidance provided by the EQUATOR Network Narrative Sections Discussion index and broader EQUATOR Network Reporting Guidelines database.
Final takeaway
A strong Discussion helps the reader understand the contribution of the study without overselling it. The section should connect findings to evidence, acknowledge uncertainty and limitations, and leave the reader with a conclusion that can be traced back to the research question and data. Once your paper is complete, learn How to Write an Academic Research Paper as a whole and enter the Manuscript Submission Process.
- How to Write the Results Section
- How to Write the Methods Section
- How to Write a Research Paper Introduction
- How to Write an Academic Research Paper
Editorial note: Reporting requirements vary by study design, discipline, journal, and jurisdiction. Always check the target journal’s current author instructions before submission.