Summary
Your findings section is where your study “speaks for itself.” Its job is to present results clearly, faithfully, and with just enough signposting that any reader—specialist or not—can follow the evidence. Start by matching your target journal’s rules (length, structure, what belongs in Findings vs Discussion). Organise around research questions or methods, use self-contained tables/figures, and write in a concise, factual past tense. Place brief, quotable “key finding” sentences at the start or end of each subsection and support them with numbers, visuals, and precise labels.
Key moves: (1) Audit the journal’s author instructions and recent exemplars; (2) plan the section with subheadings mapped to questions/hypotheses; (3) build tables/figures first and make them standalone; (4) draft in clear, objective language (effect sizes, CIs, exact p-values; or themes with representative quotations); (5) revise for accuracy, sequence, and coherence with Introduction and Discussion. Add a short end-summary that prepares readers for interpretation next.
Bottom line: report, don’t argue. Keep interpretation minimal (unless the journal combines Results/Discussion), surface the signal with tight prose and informative visuals, and verify every number twice. A transparent Findings section is the bridge that makes your Discussion credible.
📖 Full Length (Click to collapse)
How to Write the Findings Section of a Research Paper
From raw outputs to a clean, citable narrative of results
Every project is different, but excellent Findings sections share the same DNA: they are accurate, structured, minimalist in interpretation, and rich in evidence. They help readers locate the answer to each research question without hunting, and they prepare the ground for a persuasive Discussion. Below is a step-by-step playbook—equally useful for early-career researchers and for seasoned authors drafting in a new discipline.
Step 1 — Start with the Journal’s Rules and Exemplars
- Author instructions: note word limits, whether Results and Discussion are separate or combined, reporting standards (e.g., CONSORT, PRISMA, STROBE, COREQ), and any requirements for effect sizes, confidence intervals, qualitative quotes, or supplementary materials.
- Recent articles: study 2–3 papers close to your topic or method. Observe headings, sequencing, how much interpretation appears in Results, and how tables/figures are formatted and referenced.
- Scope guardrails: put background and methods elsewhere; limit interpretation unless the journal expects integrated “Results & Discussion.” Include brief comparisons to prior studies only if house style allows.
Step 2 — Plan the Architecture Around Your Questions
Map subheadings to research questions, hypotheses, or thematic codes. Use a structure that makes sense for your design:
- By research question/hypothesis: fastest for readers; each subsection ends with a “key finding” sentence.
- By method/analysis phase: useful for complex pipelines (e.g., pre-registration checks → primary analysis → sensitivity checks).
- Chronological: appropriate for longitudinal/time-series designs.
- Thematic (qualitative): group findings into major themes with subthemes and representative quotations.
Step 3 — Build Tables and Figures First (Then Write to Them)
- Self-contained visuals: captions should state what the reader is seeing, where/when, sample size, model/statistic, and one-line takeaway. Define all abbreviations in the caption or table footnotes.
- Numbering: order by first mention in text (Table 1, Figure 1, …). Never introduce a visual without explicitly directing the reader to it.
- Design: use consistent variable names, units, and decimal places; align columns; prefer en-dashes for ranges (e.g., 12–24).
Step 4 — Draft in Clear, Objective Prose
Use past tense and active voice. Keep sentences short and information-dense. Make one claim per sentence, then point to the evidence (number, table, figure).
- Quantitative: report effect sizes and 95% CIs; give exact p-values; specify model and adjustments; avoid duplicating numbers in both text and table.
- Qualitative: state themes (with definitions), provide representative quotes with attribution (pseudonym/ID, context), and indicate frequency or salience when relevant; keep analytic language light unless integrated with Discussion.
- Mixed-methods: mirror structure across strands; use joint displays when appropriate; ensure qualitative insights illuminate or explain quantitative patterns.
| Before | After (clean, citable) | Why it works |
|---|---|---|
| “There was a significant difference between groups.” | “Intervention improved recall by 8.3 points (95% CI 5.1–11.6; p<.001; Table 2).” | Specific, numeric, and anchored to a table. |
| “Participants discussed motivation a lot.” | “Theme 2—Sustained motivation: students described external scaffolds (deadlines, peer visibility). ‘Seeing others post kept me going’ (P14, female, 22).” | Named theme, definition, representative quotation. |
Step 5 — Choose the Best Sequence for Comprehension
Within each subsection, use a predictable micro-structure:
- Lead sentence: the “key finding” (one line).
- Evidence: numbers, models, or quotes; reference visuals.
- Qualification: scope, boundary conditions, sensitivity checks.
- Pointer forward: optional sentence that sets up interpretation later.
Templates You Can Copy
Key finding (quantitative):
“[Predictor] predicted [outcome] (β=value, 95% CI a–b, p=x) after adjusting for [covariates] (Figure n; Table n).”
Theme paragraph (qualitative):
“Theme 3—[Name]: participants described [definition]. This theme appeared in k/N interviews (high salience). ‘[Quote]’ (ID, context).”
Joint display sentence (mixed methods):
“Quantitative gains in Y (Δ=0.38; 95% CI 0.22–0.54) co-occurred with Theme 2—Sustained motivation, explaining adherence patterns (Table 4 shows side-by-side evidence).”
Designing Strong Tables and Figures
- Table footnotes: define abbreviations; specify model terms; denote significance with exact p-values, not asterisks alone.
- Axis labels: use plain language; avoid cryptic codes; include units.
- Colour/shape: ensure accessibility; distinguish with line types/patterns as well as colour.
Reporting Standards and Good Practice
- Statistics: avoid “p=0.000”; write “p<.001”; report test type and degrees of freedom where relevant; do not over-interpret marginal results in Findings.
- Multiple comparisons: note correction method (e.g., Holm–Bonferroni); indicate pre-registered analyses vs exploratory.
- Missing data: report extent and handling (complete case, multiple imputation, etc.).
- Qualitative trustworthiness: briefly note triangulation, coder agreement/consensus process, or member checks (details can live in Methods/Supplement).
Common Pitfalls (and Fixes)
| Pitfall | Why it hurts | Fix |
|---|---|---|
| Interpretation in Findings (when journal separates) | Blurs roles; invites reviewer pushback | Move causal claims, mechanisms, and implications to Discussion; keep neutral wording in Results |
| Duplicating the same numbers in text and tables | Wastes space; causes inconsistencies | Summarise patterns in text; reserve full values for tables |
| Ambiguous labels (var1, grpA) | Readers cannot follow or cite precisely | Use descriptive names consistently across text and visuals |
| Under-specifying models | Obscures credibility and reproducibility | State model family, link, adjustments, and key parameters |
Quantitative, Qualitative, and Mixed-Methods: Mini-Patterns
Quantitative
- Primary outcome first; secondary outcomes next; sensitivity and subgroup analyses last.
- Report effect sizes (mean difference, OR, RR, β) with CIs; include N, missingness, and model diagnostics where appropriate.
Qualitative
- Theme order reflects salience or analytic logic; each theme begins with a definition and ends with a concise synthesis line.
- Quotes are brief, vivid, and varied across participants; provide context tags.
Mixed-methods
- Parallel presentation (QUAN→QUAL) or integrated joint displays; maintain alignment between strands with cross-references.
End-of-Section Summary (Bridge to Discussion)
Close with a short synthesis that neither over-interprets nor merely repeats numbers: two to four sentences that highlight the main pattern(s), any unexpected results, and a pointer to what will be addressed in the Discussion (mechanisms, implications, limitations).
“Across analyses, intervention A improved recall and adherence, with stronger effects among first-year students. Qualitative reports of ‘sustained motivation’ align with these patterns. We next consider mechanisms and practical implications.”
Accuracy Pass: What to Check Before You Move On
- Consistency: every number matches across text, tables, figures, and supplement; totals add; percentages sum to ≈100 where expected.
- References to visuals: no orphaned tables/figures; first mention order correct; captions complete.
- Terminology: same names for variables, groups, and themes throughout.
- Line of sight: each research question is obviously answered by a subsection.
Example Micro-Makeovers
| Element | Weak | Stronger |
|---|---|---|
| Lead sentence | “We ran several analyses.” | “Attendance increased 12% (95% CI 7–17) in the intervention group (Figure 1).” |
| Theme heading | “Motivation” | “Theme 2—Sustained motivation (peer visibility as driver)” |
| Caption | “Figure 3. Results” | “Figure 3. Weekly attendance by group (n=642). Lines show means; shaded areas 95% CI.” |
Minimal Maths & Notation (When Needed)
If equations are necessary, number them and reference in text (Eq. 1). Keep derivations in an appendix; in Findings, report parameter estimates and goodness-of-fit succinctly.
Language and Style Essentials
- Prefer past tense; active voice where natural (“We observed …”).
- Avoid evaluative adjectives (“remarkable”, “striking”) in separate Results; stick to descriptors (“higher”, “lower”, “no difference”).
- Define acronyms on first use; avoid field-specific jargon if a simpler term exists.
Final Step — Revise, Read Aloud, and Align With the Rest of the Paper
- Read aloud: catch run-ons and ambiguity; shorten long sentences.
- Backward/forward alignment: ensure Findings answer the aims posed in the Introduction and supply exactly the raw material the Discussion interprets—no surprises.
- External eyes: ask a colleague (specialist) and a friend (generalist) to read for clarity and sequence; adjust headings and key sentences accordingly.
Pre-Submission Findings Checklist
- Section structured by research questions/hypotheses/themes with informative subheadings.
- Each subsection begins/ends with a one-sentence key finding.
- Tables/figures are self-contained, numbered by first mention, and referenced in text.
- Quantitative: effect sizes + CIs + exact p-values; model specification stated; missing data handling noted.
- Qualitative: themes defined; quotations attributed and varied; salience/frequency indicated when appropriate.
- No interpretation beyond what the journal permits in Results; comparisons to literature are minimal or moved to Discussion.
- Numbers consistent across text/visuals/supplement; terminology consistent throughout.
- Short closing summary bridges to Discussion.
Conclusion
A rigorous Findings section is less about eloquence and more about engineering clarity. Choose a structure that mirrors your questions, place crisp key-finding sentences where readers expect them, let well-designed visuals do heavy lifting, and keep interpretation lean until the Discussion. When you do, reviewers can verify your claims quickly, readers can cite you accurately, and your argument in the next section lands with far greater force.
