How to Write Results and Discussion Sections

Results show what the analysis found; discussion explains what those findings mean in relation to the research question, previous work and the study’s limits. Some disciplines combine the two, while others require separate chapters or sections.

The local thesis or journal guidance controls the format. Whatever the structure, readers should be able to distinguish evidence from interpretation and trace each conclusion back through the analysis to the method and data.

Return to the research questions

List the questions, hypotheses or objectives and the evidence that addresses each. This creates a complete reporting plan and helps prevent an interesting but unrelated result from taking over the chapter.

Choose an organising principle

Results may be ordered by question, hypothesis, theme, experiment, case or analytical stage. Use the same logic consistently. Chronology is useful only when sequence matters to the argument.

Report the achieved evidence base

State relevant sample, exclusions, missing material and analysis coverage. Distinguish what was planned from what was available. Do not hide an attrition or non-response pattern that changes how the results can be interpreted.

Present results without duplicating every value

Use prose to identify the important pattern and tables or figures for detailed evidence. Do not repeat every cell in a paragraph. Each visual needs a label, units, notes and a reference in the text.

Use appropriate precision

Report statistics, counts, quotations, themes or observations at the precision supported by the method. Avoid implying measurement accuracy beyond the instrument or presenting a small difference as important without the required analysis.

Handle qualitative findings transparently

Explain how themes or categories answer the question and use selected excerpts where consent and format permit. Do not treat the most vivid quotation as proof of prevalence or expose a participant through unnecessary detail.

Keep interpretation proportionate in results

A brief statement of pattern may be necessary, especially in combined sections, but save broader explanation, literature comparison and implications for the discussion where the format separates them.

Open the discussion with the answer

State the central finding in relation to the research question, then explain its significance. Avoid beginning with a second tour through every table. The discussion should move from result to interpretation.

Explain possible mechanisms

Consider why the observed pattern may have occurred and what evidence supports that explanation. Compare credible alternatives. Use causal language only when the design and analysis justify it.

Connect findings with literature

Show where the result supports, complicates or differs from relevant studies or theory. Explain meaningful differences in population, context, measure or design rather than treating disagreement as an error by one side.

Discuss unexpected findings carefully

Check data quality and analysis first, then consider plausible substantive explanations. Do not build an elaborate story around a fragile result. State when the study cannot distinguish between alternatives.

Integrate limitations at the claim level

Explain how sampling, measurement, missing data, researcher position or context affects specific interpretations. A limitation section is useful, but it should not be the first place the reader learns that a major claim is uncertain.

State implications at the right scale

Separate implications for understanding, practice, policy and future research where relevant. Keep each proposal within the evidence and authority of the study. One project rarely settles a broad debate.

End with contribution, not repetition

Summarise what the findings add and what remains unresolved. Ensure the conclusion matches the reported evidence, terminology and scope. New data or a major untested explanation does not belong in the closing paragraph.

Keep a claim-to-evidence table

Before final drafting, list each major result, its source table, figure, theme or excerpt, the interpretation you intend to make and the relevant limitation. This small audit reveals claims with no direct evidence, visuals that are never interpreted and repeated findings that can be combined.

Use cautious language precisely

Qualification should reflect a real boundary: sample, uncertainty, measurement, design or alternative explanation. Words such as may and suggests are useful only when the reader can see why caution is needed. Avoid both absolute claims and vague hedging that makes it impossible to identify the study’s actual finding.

Check consistency across the document

Verify that sample numbers, labels, themes, units and significance statements agree in the abstract, method, results, discussion and conclusion. A late revision to a table may require changes in several chapters. Search the final file for old terminology and inspect every cross-reference after export.

Ask separate questions during review

Give one reader the results and ask what they think the evidence shows before they see the discussion. Give another the discussion and ask them to locate the supporting result for each major claim. Differences reveal where presentation is unclear or interpretation has moved beyond the reported evidence.

Final checklist

• Every research question is answered by identifiable evidence.
• Figures, tables, quotations and text complement rather than duplicate each other.
• Evidence and interpretation remain distinguishable.
• Literature comparisons and limitations qualify the correct claims.
• The final contribution is stated at a scale the study can support.

Explore fields through programmes on Univs, compare listed institutes and understand the Univs study journey. Results and discussion conventions must still follow the programme, supervisor and discipline.