How to Write the Results Chapter of a Medical Thesis

Start with the analysis plan

A well-written results chapter answers each study objective in a predictable sequence. It should report what the study found, without mixing interpretation of other authors' work into every paragraph. First verify the final dataset, exclusions, missing values and denominator for each analysis.

Recommended results chapter order

1. Participant flow, recruitment and data availability.
2. Baseline demographic and clinical characteristics.
3. Descriptive measures for primary and secondary outcomes.
4. Objective-specific comparisons, associations and adjusted models where planned.
5. Safety findings, additional analyses and sensitivity checks when applicable.

Table design example

This table is fictional and demonstrates format only. Provide clearly labeled denominators if missing observations differ by variable. When presenting an effect, include confidence intervals rather than only significance flags.

Writing the accompanying narrative

The text should identify the main finding without repeating every cell. For example: “Baseline mean age was similar descriptively across the two groups; confirmatory comparisons are reported in Table 1.” If a statistically significant association is observed, state its direction and magnitude, not simply “significant difference was found.”

Quality and ethics checks

• Ensure tables and charts derive from the same verified dataset.
• Describe missing observations and changes in the analysis population.
• Use p < 0.001 rather than p = 0.000.
• Do not selectively omit non-significant results specified in the protocol.
• Do not claim that AI-generated narrative is independently validated.

For observational studies, review the relevant STROBE checklist. For randomized trials, consult CONSORT 2025.

The existing Statistical Analysis tool can help prepare tables, graphs and draft result explanations for review.

Frequently asked questions

Should I interpret earlier studies in Results?

Generally keep primary interpretation and comparison with prior literature in Discussion, while presenting observed estimates, uncertainty and planned analysis outputs in Results.

How should missing data be reported?

State denominators, missing counts and any handling method clearly for each relevant analysis.

Further reading and primary guidance

• ICMJE manuscript preparation: statistics, tables and references
• EQUATOR Network: STROBE observational study reporting guideline
• SPIRIT 2025 and CONSORT 2025 statements