How to Write a Discussion Section

The discussion is where your results become knowledge. It is also the section most often marked down, because many students restate findings instead of interpreting them.

Your results section tells the reader what you found. The discussion tells them what it means, how it fits with what was already known, and how much weight it can bear. It is the part of a research paper or dissertation that best shows your thinking, and markers read it closely. The structure below works for most empirical projects across disciplines.

Open by answering the research question

Begin the discussion with a short paragraph that directly answers your research question or states whether your hypotheses were supported. Do not make the reader wait. One or two sentences summarising the principal findings orients everything that follows.

Interpret, do not repeat

A common fault is a discussion that reruns the results with added adjectives. You do not need to restate statistics or quotes in detail; refer to them briefly and move to interpretation. For each key finding, ask what it suggests, why it might have occurred and what it implies. Interpretation is the core of the section.

Compare findings with existing research

Place each main finding in the context of the literature you reviewed. Where your results agree with previous studies, say so and consider what your study adds — a new population, setting or method. Where they disagree, explore possible reasons: differences in sample, measures, context or design. Disagreement handled well is often more interesting to markers than agreement.

The discussion should mainly draw on sources already introduced in your literature review. Bringing in many new ones suggests the review was incomplete.

Explain unexpected results

Surprising or null findings are not failures to hide. Offer plausible explanations, grounded in theory or in features of your study, and be honest about which explanation seems most likely. Avoid overconfident claims; phrases such as “one possible explanation” and “this may reflect” signal appropriate caution.

Link back to theory

If your study used a theoretical framework, return to it. Do your findings support it, extend it or challenge part of it? Showing how empirical results speak to theory is what lifts a discussion from description to scholarship, and it is often the difference between a good mark and an excellent one at master’s level.

Consider alternative explanations

Strong discussions do not only offer the most convenient interpretation of a finding. They consider other explanations and weigh them. Could the result reflect a confounding variable, a measurement issue, the characteristics of your sample or the timing of data collection? Addressing these openly shows critical thinking and anticipates the questions an examiner would ask. Where you cannot rule an alternative out, say so, and explain what evidence would distinguish between the explanations.

State the implications

Explain who should care about your findings and why. Implications might be practical — for teachers, clinicians, managers or policymakers — or theoretical and methodological, for other researchers. Keep them proportionate to your evidence. A small, single-site study can suggest a promising direction; it cannot justify a national policy change.

Acknowledge limitations honestly

Every study has limitations, and examiners expect you to identify the important ones yourself. Focus on those that genuinely affect interpretation — sample size or composition, measurement, design, response rate, researcher influence — and explain how each might have shaped your results. Avoid generic lists, and avoid undermining your entire study: limitations qualify findings, they do not erase them.

Note strengths as well

A balanced account also recognises what the study did well: a robust design, a hard-to-reach sample, a validated instrument or rich data. Briefly stating strengths alongside limitations helps the reader judge how much confidence your conclusions deserve.

Suggest specific future research

Recommendations for future research should follow from your findings and limitations, not be generic. “More research is needed” tells the reader nothing. A better recommendation names what should be studied, with whom and how — for example, a longitudinal design to test whether an association you observed holds over time.

Get the tone right

The discussion requires a measured, confident voice. Avoid both extremes: overclaiming (“this proves”) and excessive hedging, where every sentence is qualified until it says nothing. Use language that matches the strength of your evidence — “suggests,” “is consistent with,” “provides strong evidence for.” Write in the past tense for your own findings and the present tense for established knowledge and for what your findings imply.

Keep the structure clear

Organise the discussion around your main findings or research questions, using subheadings in longer documents. Each paragraph should tackle one idea: the finding, its interpretation, the comparison with literature and the implication. Some disciplines combine discussion and conclusion; others separate them. Check your guidelines and, if separate, keep the final summary for the conclusion chapter.

Common mistakes to avoid

• Restating results in detail instead of interpreting them.
• Introducing new findings that did not appear in the results section.
• Ignoring results that do not fit the expected pattern.
• Comparing with literature in a list (“Smith found… Jones found…”) rather than in an argument.
• Generic limitations and future research that could apply to any study.