How to Write the Discussion Section

The Discussion interprets your results: it states the main finding, compares it with other studies, explains it, acknowledges limitations and describes implications. It should answer the question you posed in the Introduction and not repeat the Results.

A reliable structure

1. Main finding: one or two sentences answering the objective.
2. Comparison with existing evidence: where it agrees or differs, and possible reasons.
3. Interpretation: plausible explanations, labelled as such if untested.
4. Strengths and limitations: be specific about bias, confounding, sample, measurement, generalisability.
5. Implications: for clinical practice, policy or future research.
6. Conclusion: a short statement supported by the data (see How to Write a Strong Scientific Conclusion).

Writing tips

• Do not introduce new results.
• Use cautious wording for observational findings ("was associated with", not "caused").
• Be balanced: discuss findings that do not fit your hypothesis.
• Keep to the question: a long literature review belongs elsewhere.

Illustrative example

Invented for teaching. "Surgery within 48 hours was associated with lower delirium. This agrees with two cohort studies but not with one trial that included younger patients. Residual confounding by frailty is possible, because frailty was not recorded in our data. These results suggest that timing may matter and support a randomised trial." (Invented example.)

Writing about disagreement with earlier studies

If your result differs from earlier work, compare design, population, measurement and analysis before suggesting why. Differences may reflect real variation or bias in either study. Do not dismiss earlier findings because they conflict with yours, and do not adopt them uncritically. Offer explanations as possibilities and say what further evidence would settle the question.

Common pitfalls

• Restating Results.
• Overgeneralising.
• Hiding limitations or listing only generic ones.

Key takeaways

• Answer the question, compare, qualify and conclude.
• Be specific about limitations and implications.

This page is general educational information about research methods. It is not medical, legal or regulatory advice; follow your institution's ethics board and your target journal's instructions.