How to Write a Discussion Section: Explain What Your Results Mean

The discussion section is where a paper becomes more than a report of numbers, quotations, or observations. Your job is to answer the reader’s quiet question: so what? Return to the research question, explain the pattern you found, and show why that pattern matters in the context of earlier work.

Begin with the answer, not a repeated results table

Open with two or three sentences that state the main finding in plain language. Do not copy the results section or list every statistic again. A useful opening names the pattern, tells the reader whether it supports the expectation, and points back to the question that guided the study.

For example, instead of writing “Table 2 shows a difference,” write “Students who used retrieval practice recalled more concepts after one week, supporting the prediction that active recall improves retention.” The exact evidence remains important, but interpretation should lead.

Build each paragraph around a claim

A reliable discussion paragraph has four moves: make an interpretive claim, point to the relevant result, compare it with a source or theory, and explain the implication. This structure prevents a common problem: jumping from one interesting observation to another without an argument.

When studies disagree, do not force agreement. Consider differences in sample, setting, measurement, or timing. A thoughtful explanation of a difference is stronger than pretending it does not exist. If source-heavy sentences become difficult to follow, use the Academic Proofreader to check clarity and sentence flow.

Separate explanation from speculation

You may propose reasons for a result, but label uncertainty honestly. Words such as “may,” “suggests,” and “is consistent with” are not weak when the evidence is limited; they are accurate. Avoid claiming causation from a correlational design, and do not introduce data that the study did not collect.

Treat limitations as analysis

A limitations paragraph should do more than apologize. Name the constraint, explain how it could shape the result, and suggest a specific next step. A small convenience sample, for instance, may limit generalization but still reveal a useful pattern worth testing with a broader group.

End by returning to the contribution

Close with the contribution your study can reasonably make: a refined explanation, a practical recommendation, a question for future research, or evidence that challenges an assumption. Before submission, use the Tone Refiner to make cautious claims precise, then run the Originality Scanner on paraphrased research passages. The final discussion should sound like your reasoned interpretation—not a summary pasted from the results.

A final checklist

Ask: Did I answer the research question? Did every major claim point to evidence? Did I distinguish findings from possible explanations? Did I state a meaningful limitation and a realistic implication? If the answer is yes, your discussion is doing its most important work.