Key Takeaways
- The results section presents your findings without interpretation. LLMs can help you describe data clearly, structure the section, and write around tables and figures.
- Never let an LLM interpret your results. It does not know your study, your data, or your research context. Interpretation belongs in the discussion section, written by you.
- The best use of LLMs for results writing is clarity: turning raw statistical output into readable prose, describing table and figure content, and checking for consistency between text and data.
- These 10 prompts cover describing statistical results, writing around tables, structuring results by research question, reporting qualitative themes, and pre-submission checks.
- For a comprehensive guide to using LLMs across all stages of research, see our LLM Prompting Guide for Researchers.
The results section is the factual backbone of your paper. It presents what you found, supported by tables, figures, and statistical values. The writing should be precise, clear, and free of interpretation. That makes it a section where LLMs can be genuinely useful: they can help you describe your data in well-structured prose without the temptation to add commentary.
The risk is that an LLM might add interpretation where it does not belong, or describe findings inaccurately. Every output must be verified against your actual data. For broader LLM guidance, see our LLM Prompting Guide for Researchers.
The 10 Prompts
Prompt 1: Describe Statistical Results in Prose
Convert the following statistical output into academic prose suitable for a results section. Report the test used, test statistic, degrees of freedom, p-value, and effect size. Use past tense. Do not interpret the results or discuss their implications. Just describe what was found.
[paste your statistical output, e.g., “t(198) = 3.42, p = .001, d = 0.48”]
Prompt 2: Write Text Around a Table
The following table presents [describe what the table shows]. Write 2-3 sentences of results text that references this table, highlights the most important findings, and guides the reader’s attention to the key rows or columns. Do not repeat every number in the table. Just narrate the main takeaways. Use the format “as shown in Table X.”
[paste or describe your table data]
Prompt 3: Structure Results by Research Question
My paper has the following research questions:
1. [RQ1]
2. [RQ2]
3. [RQ3]
And the following key findings:
[list your findings]
Organise these findings into a results section structure with subheadings that correspond to each research question. For each subsection, suggest what tables/figures should be referenced and what statistical information should be reported.
Prompt 4: Report Qualitative Themes
I conducted qualitative research with [number] participants. My thematic analysis produced the following themes:
Theme 1: [name] – [brief description]
Theme 2: [name] – [brief description]
Theme 3: [name] – [brief description]
For each theme, write an introductory paragraph that: names the theme, explains what it captures, states how many participants referenced it, and sets up the supporting quotes I will add. Use past tense and academic tone.
Prompt 5: Describe a Figure or Chart
I have a [type: bar chart / line graph / scatter plot / etc.] that shows [what it displays]. Write 2-3 sentences describing the key trends, patterns, or comparisons visible in the figure. Reference it as “Figure X.” Do not interpret why the pattern exists. Just describe what the figure shows.
Key data points: [list the main values or trends]
Prompt 6: Write a Demographic Summary
Write a participant demographics paragraph for the results section using the following data. Report it in academic prose (not a list). Include percentages and means where appropriate. Reference “Table 1” for the full breakdown.
Total participants: [N]
Gender: [breakdown]
Age: [mean, SD, range]
Education: [breakdown]
Other relevant demographics: [list]
Prompt 7: Check for Interpretation in Results
Review the following results section and flag any sentences that contain interpretation, explanation, or comparison to previous literature. The results section should only describe what was found, not why. For each flagged sentence, suggest whether it should be removed, moved to the discussion, or rewritten as a purely descriptive statement.
[paste your results section]
Prompt 8: Report Non-Significant Results
The following analysis produced a non-significant result: [describe the test and values]. Write 1-2 sentences reporting this result in academic prose. Use neutral language. Do not imply that the result is unimportant or that the hypothesis was “wrong.” Simply report the finding: no significant difference/relationship was found, with the statistical values.
Prompt 9: Generate a Table Caption
Write a concise, informative table caption for Table [number]. The table presents [what it shows: e.g., “descriptive statistics for all study variables” or “regression coefficients for Model 1”]. The caption should be one sentence, begin with “Table X.”, describe the content clearly, and include the sample size if relevant. Follow APA 7th edition format.
Prompt 10: Cross-Check Text Against Data
Compare the following results text against the data in the table below. Flag any inconsistencies: numbers that do not match, findings described in the text but missing from the table, or table data not mentioned in the text. List each inconsistency with the specific values involved.
Results text: [paste]
Table data: [paste or describe]
Accuracy Is Everything in the Results Section
Every number, every statistical value, and every description of a table or figure must match your actual data. LLMs can help you write clearly, but they cannot verify your data. After using any prompt above, cross-check every number in the output against your original analysis. A single misreported p-value or effect size can undermine reviewer confidence in your entire paper.
For guidance on structuring the full paper, see our post on how to structure and organise a research paper.
Need your results section professionally checked? PM Proofreading’s journal editing service ensures your results are clearly presented, properly formatted, and consistent with your tables and figures. Upload your manuscript and get a quote.
Further Reading
- LLM Prompting Guide for Researchers: ChatGPT, Claude, and Gemini
- How to Structure and Organize a Research Paper
- Qualitative vs Quantitative Research
- How to Write the Perfect Research Abstract
Frequently Asked Questions
Can I use AI to analyse my data?
LLMs are not statistical tools. They can help you write about your results, but they should not be used to run statistical analyses. Use dedicated software (SPSS, R, Stata, Python) for analysis. Use LLMs to convert raw output into readable academic prose after the analysis is complete.
Should interpretation appear in the results section?
No. The results section describes what you found. The discussion section explains what it means. Keep interpretation, comparison to previous studies, and implications out of the results section. Use Prompt 7 to check for interpretation that has crept in.
How do I report non-significant findings?
Report them the same way you report significant findings: state the test, the test statistic, the p-value, and the effect size. Use neutral language. “No significant difference was found between groups, t(198) = 1.24, p = .216, d = 0.18.” Non-significant results are valid findings that contribute to the evidence base.
How much of the table data should I repeat in the text?
Highlight the key findings only. Do not repeat every number from the table in the text. Reference the table (“as shown in Table 3”) and draw the reader’s attention to the most important rows, columns, or comparisons. The text narrates; the table provides the detail.
