Why Your Results Section Matters
You’ve completed the research. You’ve run the analyses. Now you need to write the results section—the part of your paper where you present what you actually found. This is where many students and even experienced researchers stumble. They either interpret their findings prematurely, report statistics incorrectly, or fail to distinguish results from discussion. In quantitative research, the results section is your evidence. It’s what allows readers to evaluate your work, replicate your findings, and trust your conclusions. Done poorly, it invalidates your entire study. Done well, it provides a clear, objective foundation for your contribution to the field.
The purpose of data analysis writing in the results section is singular: to report what the data show, nothing more. No interpretation. No speculation about why. Just the facts, formatted precisely according to your target journal’s or instructor’s style guide. This separation is fundamental to scientific communication—the results tell us what happened; the discussion explains what it means.
- How to structure a results section that reports findings without interpretation
- The critical differences between APA, AMA, and Chicago formatting for statistics
- Why exact p-values matter more than p < .05
- Common mistakes that undermine credibility (and how to fix them)
- Discipline-specific considerations for sciences, social sciences, medicine, and engineering
- Complete worked example from hypothesis to polished results paragraph
This guide covers everything you need to write an effective statistical results section. We’ll examine the universal structure that works across disciplines, detail the formatting requirements of the three major style guides (APA, AMA, Chicago), highlight common mistakes that even seasoned researchers make, and provide discipline-specific guidance. You’ll learn how to integrate tables and figures effectively, use the right tense and voice, and format statistical symbols correctly. Most importantly, you’ll understand the principle that makes results sections readable, credible, and publishable: objectivity through precise reporting.
Understanding the Results Section: What It Is (And Isn’t)
Definition and Purpose
The results section is the narrative account of your statistical findings. It presents the data you collected, the analyses you performed, and the outcomes of those analyses—organized logically, formatted correctly, and stated objectively. According to the American Psychological Association, the results section should “present the findings of your study without interpretation, in a neutral and objective manner” (Purdue OWL, 2023). This objectivity is not optional; it’s the foundation of empirical research.
The results section answers specific research questions through statistical tests. If your hypothesis was “Group A will score higher than Group B on the memory test,” your results section states whether that’s true, supported by the actual t-test or ANOVA output. It reports: “The difference was significant, t(58) = 2.45, p = .017.” Full stop. The discussion later explains why that difference exists and what it means for memory theory.
Key Distinction: Results Report WHAT; Discussion Explains WHY
This distinction cannot be overstated. Results = what you found. Discussion = why it matters. The results section tells us that “treatment group participants lost an average of 5.2 kg (SD = 1.8) compared to 0.7 kg (SD = 1.2) in the control group, F(1, 84) = 12.34, p = .001.” The discussion interprets that finding: “The 5.2 kg average loss suggests our intervention is clinically meaningful, possibly because it combined dietary changes with behavioral support, addressing both caloric intake and habit formation.”
Mixing these two purposes creates what reviewers call “results-discussion confusion,” a fatal flaw that leads to desk rejection or poor grades. When you find yourself writing “This likely occurred because…” or “These findings suggest that…”—stop. That’s discussion. Move it to the discussion section. The results section is purely descriptive.
When Results and Discussion Can Be Combined
Some formats—particularly theses, dissertations, and certain journal articles (especially in medicine and engineering)—combine results and discussion into a single “Results and Discussion” section. This format is common when the study is exploratory, when space is limited, or when the discipline expects integrated presentation. Even then, you should clearly separate the two components within the combined section. A common approach: present the statistical finding first, then immediately follow with a brief interpretation sentence. But use this format only when explicitly required; most academic journals and APA-style papers demand separate sections.
Past Tense vs. Present Tense
Results are what you found. Therefore, use past tense: “The analysis showed…” “Group A performed better…” “The correlation was significant…” The only exception is when referring to figures or tables that exist in the present document: “As shown in Table 1…” “Figure 2 illustrates…” These use present tense because the table and figure are current elements of your paper.
Active vs. Passive Voice
Traditionally, scientific writing favored passive voice (“It was found that…”) to maintain objectivity and de-emphasize the researcher. Modern style guides increasingly accept active voice (“We found that…”) because it’s clearer and more direct. APA 7th edition explicitly states that authors “should use first person to refer to themselves and their work rather than third person” when appropriate (APA, 2020). The key is consistency and clarity. If your discipline or journal uses passive voice, follow that convention. When in doubt, active voice is usually acceptable and often preferred for readability.
The Universal Structure of an Effective Results Section
Regardless of discipline or statistical method, an effective results section follows a logical order that mirrors your research process. This structure makes your findings easy to follow and demonstrates methodological transparency.
Typical Order: Descriptive → Inferential → Supplemental
Start with descriptive statistics—means, standard deviations, frequencies, medians, interquartile ranges. These tell the reader what your sample looked like and provide baseline information. For each variable of interest, report its central tendency and variability. If you have groups, report these separately.
Next, present inferential statistics—the tests that answered your research questions. Organize these by hypothesis or research question, not by statistical method. If you had three hypotheses, address them in order. For each, state the test used, the test statistic, degrees of freedom, p-value, and effect size. Reference tables or figures for detailed data.
Finally, report supplemental analyses if you conducted any: post-hoc tests, subgroup analyses, assumption checks, exploratory analyses. Be transparent about what was planned versus exploratory.
Organizing by Research Question or Hypothesis
Don’t organize your results by statistical test (e.g., “t-tests, then ANOVAs, then regressions”). That makes readers work to connect findings to your research questions. Instead, structure results around your hypotheses:
H1: Participants receiving intervention will show greater improvement than controls.
Results: The intervention group improved significantly more (M = 12.5, SD = 3.2) than controls (M = 8.1, SD = 2.9), t(58) = 4.21, p < .001, d = 0.75.
H2: The intervention effect will be stronger for participants with high baseline anxiety.
Results: …
This structure is reader-friendly and shows you tested what you said you would test.
Subheadings for Different Types of Analyses
Use descriptive subheadings to guide readers. For example:
- Participant Characteristics
- Primary Outcome Analysis
- Secondary Outcomes
- Post-Hoc Comparisons
- Exploratory Findings
These subheadings are not mandatory in all journals (some use only H2 for major sections), but they dramatically improve readability, especially in longer papers or theses.
What to Include: All Relevant Findings
Report all results that address your research questions—including non-significant findings. Omitting non-significant results is unethical and introduces publication bias. If you tested it and it’s relevant to your hypotheses, report it. “The difference between groups was not statistically significant, t(42) = 0.83, p = .41, d = 0.13.” This transparency is essential for scientific integrity.
What to Exclude: Interpretation, Raw Data, Methodology Repetition
- Interpretation: Move “why” explanations to discussion.
- Raw data: Individual participant scores do not belong in the results section (unless you’re publishing a data descriptor paper). Use tables for summarized data; put full datasets in supplements or appendices.
- Methodology repetition: Don’t re-explain your procedures. Briefly remind readers if necessary (“As described in the Methods, we conducted a mixed ANOVA…”), but don’t repeat detailed methods. Assume readers have read the methods section.
Reporting Statistical Results: The Big Three Style Guides
Different disciplines use different style guides, each with specific formatting rules for statistics. Using the wrong format signals inexperience and may lead to point deductions or manuscript rejection. Here’s what you need to know about the three most common guides.
APA 7th Edition (Social Sciences, Psychology, Education)
APA style is the standard for psychology, education, and many social sciences. It’s precise about statistical formatting.
Basic format: Test statistic (italicized), degrees of freedom, exact p-value, effect size.
t(50) = 2.45, p = .034, d = 0.50
F(2, 87) = 3.56, p = .032, η² = 0.08
r(48) = 0.47, p = .001
χ²(3) = 9.82, p = .020
Key rules:
- Italicize statistical symbols: p, t, F, r, χ², z, N, M, SD, SE, CI. Do not italicize df or ns.
- P-values: Report exact values (p = .042, not p < .05). Round to two or three decimal places as appropriate. For p < .001, write “p < .001”.
- Leading zeros: Omit the zero before the decimal for p, d, r, η², and other probabilities and correlations. Use the zero for t, F, M, SD (e.g., t = 2.45, M = 12.5, SD = 3.2).
- Effect sizes: Always report effect sizes with inferential tests. Common ones:
- Cohen’s d: d = 0.2 (small), 0.5 (medium), 0.8 (large)
- Pearson’s r: r = 0.1 (small), 0.3 (medium), 0.5 (large)
- η² (eta-squared): η² = 0.01 (small), 0.06 (medium), 0.14 (large)
- Confidence intervals: Report 95% CIs for means, mean differences, and effect sizes: M = 12.5, 95% CI [10.2, 14.8].
- Tables and figures: Tables have titles in italics, title case, above the table. Figures have captions below, in sentence case. Both must be referenced in text.
Example paragraph:
Participants who received cognitive-behavioral therapy reported significantly lower anxiety scores (M = 18.3, SD = 5.2) than those in the waitlist control group (M = 24.7, SD = 4.9), t(58) = 4.23, p < .001, d = 0.78. The 95% CI for the mean difference was [-9.1, -3.4].
AMA Style (Medicine, Clinical Research)
AMA style, used in medical journals, emphasizes clinical significance and confidence intervals.
Basic format: Italicize P, t, F, r, N. Report to 2 decimals (P to 3 if < .01). Use en dash for CI ranges.
P = .03
OR, 1.25 (95% CI, 1.10-1.40)
t = 2.45, P = .017
mean (SD) or median (IQR)
Key rules:
- P values: Capital P, italicized. Report exact values to three decimal places if P < .01; otherwise two decimals (P = .03). For very small P values, write P < .001.
- N vs n: Capital italic N = total sample size in study. Lowercase italic n = subgroup size.
- Means and variability: Report as mean (SD) or median (interquartile range [IQR]). Example: 12.5 (3.2) or 18 (15-22).
- Odds ratios and risk: Report with 95% CI: OR, 1.25 (95% CI, 1.10-1.40); RR, 0.67 (95% CI, 0.52-0.86).
- Confidence intervals: Use en dash (–) between limits, not hyphen or “to”: 95% CI, 1.10-1.40.
- Percentages: Use the % symbol with numbers: 45% (not forty-five percent).
- Tables and figures: Table titles above, in sentence case. Figure legends below. Include explanations of symbols, abbreviations, and statistical significance indicators.
Example paragraph:
The intervention group achieved a mean reduction in systolic blood pressure of 12.5 mm Hg (SD = 4.2) compared to 2.1 mm Hg (SD = 3.8) in controls (t = 7.32, P < .001). The odds of achieving target blood pressure were 2.3 times higher in the intervention group (OR, 2.3; 95% CI, 1.5-3.6).
Chicago Author-Date (Social Sciences, History, Sciences)
Chicago Author-Date, used in many social sciences and some natural sciences, emphasizes in-text citations with year.
In-text citations: (Author Year, page) or Author (Year).
Reference list: Author. Year. Title. Publisher. URL.
For statistics themselves, Chicago doesn’t have as many rigid formatting rules as APA or AMA, but follows general scientific conventions:
- Italicize statistical symbols: t, F, r, p, N, M, SD
- Report p-values to two or three decimals
- Use consistent decimal places throughout
- Include test statistics, df, and p-values
- % sign with numbers
Example:
As shown in Table 1, treatment effects were substantial (M = 15.3, SD = 2.1 for treatment; M = 10.2, SD = 1.9 for control), t(42) = 6.81, p < .001 (Smith 2023, 45).
The main Chicago requirement is consistent citation of sources in the reference list, not the specific formatting of statistical values per se. However, always check the specific journal or instructor requirements, as some Chicago-based publications have their own statistical formatting rules.
Essential Elements: What Every Results Paragraph Needs
Every paragraph in your results section should contain certain core elements, arranged in a logical sequence. This template works for virtually any statistical result.
1. Restate Hypothesis or Research Question Briefly
Don’t assume readers remember what you were testing. Remind them:
To test our first hypothesis that social support would buffer stress, we compared cortisol levels…
2. Present Descriptive Statistics First
Before inferential tests, readers need to know the baseline characteristics:
The high-stress group had a mean cortisol level of 18.4 µg/dL (SD = 4.2, n = 32) compared to 12.1 µg/dL (SD = 3.8, n = 30) in the low-stress group.
For group comparisons, report descriptive stats for each group. For correlations or regressions, report means and SDs of the variables.
3. Report Inferential Statistics with All Components
For each hypothesis test, include:
- The test used (t-test, ANOVA, regression, χ²)
- Test statistic (italicized)
- Degrees of freedom
- Exact p-value (or p < .001)
- Effect size (Cohen’s d, r, η², OR, etc.)
Example:
The difference was statistically significant, t(60) = 3.84, p = .001, d = 0.66.
4. Include Effect Sizes for Practical Significance
P-values tell you whether an effect exists; effect sizes tell you how big it is. A finding can be statistically significant but practically meaningless (e.g., p = .049 with d = 0.1). Always report effect sizes. As the NCBI guidelines emphasize, “statistical significance is not synonymous with practical significance” (Harhay et al., 2020).
5. Mention Direction of Effects
State whether the effect was positive or negative, higher or lower, increase or decrease. Don’t make readers infer direction from the table:
- Group A scored higher (not just “Group A differed from Group B”)
- The correlation was positive
- Symptoms decreased over time
6. Reference Tables and Figures Appropriately
If you have a table with comprehensive statistics, your text can be a summary:
As shown in Table 1, the treatment group showed significantly greater improvement across all outcomes. For primary outcome scores, the treatment group mean was 15.2 (SD = 3.4) compared to 22.8 (SD = 4.1) in controls, a difference of 7.6 points (95% CI, 5.9-9.3; t = 7.45, p < .001).
Rule: Every table and figure must be mentioned in the text, in the order they appear.
7. Use Plain English to Bridge Statistics
Don’t just dump numbers. Use narrative to connect:
Weak: “t(58) = 2.45, p = .017.”
Better: “The treatment group scored significantly higher than controls, t(58) = 2.45, p = .017, indicating that the intervention had a measurable effect.”
But don’t interpret—don’t say “this suggests that…” or “likely due to…”
Common Mistakes (And How to Avoid Them)
Even experienced researchers make these errors. Here’s what to watch for.
| Mistake |
Why It’s Wrong |
How to Fix |
| Interpreting findings |
Results should report, not explain |
Save interpretation for discussion; state findings objectively |
| Reporting p < .05 |
Vagueness; modern standards require exact p |
Report exact p = .042; if p < .001, write “p < .001” |
| Including raw data |
Results section is for summarized findings |
Put raw data in appendix/supplement; use tables for summaries |
| Repeating table values in text |
Redundant; wastes space |
Highlight key trends in text; let tables contain details |
| Using present tense for findings |
Results are completed actions |
Use past tense: “showed,” “increased,” “was significant” |
| Ignoring non-significant results |
Biased reporting; transparency required |
Report all relevant findings, including ns results |
| Mismatched numbers (text vs table) |
Confuses readers; looks careless |
Double-check all values match exactly between text and tables |
| Missing effect sizes |
Statistical ≠ practical significance |
Always include effect size (Cohen’s d, r, η², OR) with p-values |
| Wrong italics/formatting |
Violates style guide; appears unprofessional |
Review APA/AMA/Chicago rules; italicize p, t, F, r, M, SD but not ns |
| Incorrect decimal places |
Inconsistent precision looks sloppy |
Follow style guide: usually 2 decimals for means, exact p-values to 3 decimals if needed |
Integrating Tables and Figures Effectively
Tables and figures are not decorations—they’re integral to your results presentation. Used properly, they make dense information scannable. Used poorly, they confuse readers.
Placement: Immediately After First Mention
In most formats, tables and figures appear either:
- Immediately after the paragraph where they’re first mentioned, OR
- At the end of the manuscript in a separate “Tables and Figures” section (some journals)
Check your target format’s requirements. When in doubt, place them immediately after first reference in text.
Text Role: Summarize Trends, Don’t Describe Every Value
The text should highlight patterns, comparisons, and notable values. The table contains the details.
- Ineffective: “Group A had a mean of 15.2, SD 2.1; Group B had mean 18.7, SD 2.5; Group C had mean 22.1, SD 3.0…”
- Effective: “As shown in Table 2, mean scores increased across groups in a dose-response pattern, with the highest dose group showing the greatest improvement.”
Caption Rules: Figures Below, Tables Above
- Table caption: Above the table, in title case (APA) or sentence case (AMA), italicized (APA). Include table number and descriptive title.
- Figure caption: Below the figure, in sentence case. Include figure number, descriptive title, and explanation of symbols/abbreviations.
Example table caption (APA):
Table 1
Means and Standard Deviations for Outcome Variables by Treatment Group
Example figure caption (AMA):
Figure 1. Survival curves for treatment vs. control groups. Tick marks indicate censored observations. HR = hazard ratio; CI = confidence interval.
Caption Content: Number, Title, Explanation
Good captions are self-contained. A reader should understand the table/figure without reading the text. Include:
- What’s being shown
- What the symbols/abbreviations mean
- Any important notes (e.g., “Error bars represent ±1 SE”)
Self-Explanatory: Understandable Without Text
If readers have to flip back to the methods to understand your table, it’s not good. Include enough information in captions or table headings to make it standalone. Define all abbreviations the first time they appear.
Consistency: Same Formatting Throughout
- Same number of decimal places for similar measures
- Consistent order of columns
- Same font and spacing
- Uniform symbol definitions
Referencing: “As Shown in…” or “Table X Displays…”
Always reference tables and figures in the text:
- “The demographic characteristics are presented in Table 3.”
- “As illustrated in Figure 1, survival was significantly longer…”
Don’t just place a table without mentioning it.
Discipline-Specific Considerations
The general principles above apply everywhere, but disciplines have their own conventions. Adapt accordingly.
Sciences (Biology, Chemistry, Physics)
Sciences often prefer concise presentation with heavy use of figures. Many scientific journals combine results and discussion, though this varies. Biology frequently uses figures for trends and reserves tables for numerical data with multiple variables. Chemistry may present spectral data or reaction yields. Physics emphasizes uncertainty reporting (± values) and significant figures.
Key science conventions:
- Figures should be high-resolution (300+ dpi for print)
- Error bars on graphs
- Uncertainty values reported with measurements
- Sometimes active voice is more common: “We observed that…”
- Replication studies may have separate “Validation” subsections
Social Sciences (Psychology, Sociology, Education)
Social sciences typically require strict separation of results and discussion, following APA 7th edition. Results are purely descriptive; interpretation begins in the discussion. Effect sizes are increasingly emphasized alongside p-values. Tables are common for group comparisons; figures for interaction effects or trends over time. Qualitative results are reported separately, often in thematic narratives rather than statistical tables.
Key social science conventions:
- No interpretation in results
- Detailed statistical reporting (test statistics, df, p, effect size)
- Confidence intervals for estimates
- Power analysis often reported in methods, not results
- Transparency about missing data and exclusions
Medicine/Clinical Research
Medical writing follows AMA Manual of Style. Clinical significance (not just statistical) is paramount. P-values less emphasized than confidence intervals and effect sizes. Survival analysis (Kaplan-Meier curves) is common. Tables often show baseline characteristics, primary outcomes, adverse events. CONSORT guidelines govern randomized trial reporting. Odds ratios, relative risks, and number-needed-to-treat are standard effect measures.
Key medical conventions:
- Report both absolute and relative effects when relevant
- Include number at risk in survival curves
- Adverse events must be reported
- Intention-to-treat analysis preferred
- Clinicaltrials.gov registration numbers required for trials
Humanities with Quantitative Components
Humanities disciplines that incorporate quantitative methods (history, literary analysis with text mining, archaeology) often use Chicago Author-Date. The writing may be more narrative, with statistics integrated into prose rather than isolated in tables. Statistical reporting is generally less dense than in sciences—fewer tests, more emphasis on what the numbers mean for humanistic inquiry. Still, basic formatting rules (italics for symbols, proper p-value reporting) apply.
Engineering/Computer Science
Engineering and CS have varied practices. Conference papers often combine results and discussion due to page limits. Journal articles tend to separate them. Performance metrics (accuracy, F1-score, execution time, memory usage) are reported in tables. Statistical tests are reported when comparing algorithms. Many engineering journals have their own style guides that modify APA or Chicago.
Key engineering conventions:
- Algorithm performance tables with means and standard errors across benchmark datasets
- Statistical significance indicated with asterisks (*p < .05, **p < .01) in tables
- Box plots and bar graphs common
- May use combined results/discussion format
Writing Style: Tense, Voice, and Tone
Tense: Past for Findings, Present for Figures/Tables
Standard rule: Past tense for your statistical results because they are completed actions:
“Participants who exercised showed greater improvement (M = 8.2, SD = 1.5) than sedentary participants (M = 5.1, SD = 1.3), t(45) = 4.67, p < .001.”
Present tense only for:
- Figures/tables that exist in your current document: “As shown in Table 1…”
- General statements about what your paper does: “This section presents…”
Voice: Active Increasingly Acceptable
Traditional scientific writing mandated passive voice to maintain objectivity: “It was found that the treatment was effective.” Modern style guides (APA 7th, AMA 11th) accept—and sometimes prefer—active voice for clarity:
“We found that the treatment was effective.”
“The analysis showed…”
Active voice is more direct and uses fewer words. When in doubt, follow your discipline’s conventions. Psychology and medicine generally accept active voice. Some traditional fields (certain areas of physics, chemistry) may still prefer passive. Check recent articles in your target journal.
Objectivity: Third-Person, No “I” or “We” When Appropriate
Even with active voice, maintain scientific objectivity. In some contexts, especially when the focus should be on the research rather than the researchers, use third person without pronouns:
“The treatment group showed improvement” (preferred over “We found that the treatment group showed improvement”)
But APA 7th explicitly states that using first person plural (“we”) is acceptable when referring to the research team. The key is consistency and appropriateness for your context.
Hedging: Minimal in Results (Unlike Discussion)
The results section is not the place for hedging or qualifying language. State what the data show:
Avoid: “The results might suggest a trend toward improvement.”
Use: “The treatment group improved significantly more than controls.”
Hedging belongs in the discussion when you’re speculating beyond the data.
Precision: Exact Numbers, Correct Decimals, Consistent Formatting
Precision matters. Report numbers with appropriate significant figures:
- Means and standard deviations: usually one decimal more than raw data
- P-values: 2-3 decimals as needed; exact, not ranges
- Test statistics: usually two decimals
- Effect sizes: two decimals typically
Be consistent. If you report one mean to one decimal (12.5), report all means to one decimal (18.3, not 18.27). Follow your style guide’s specific rules.
Clarity: Explain Complex Statistical Concepts Alongside Technical Terms
Your readers may not be statisticians. When using technical terms, briefly explain:
“Cohen’s d (a measure of effect size) of 0.78 indicates a large practical effect.”
But don’t over-explain basic terms like mean, standard deviation, or p-value. Assume college-level statistics knowledge.
A Complete Example: From Hypothesis to Publication
Let’s walk through a realistic example from hypothesis to finished results paragraph, showing proper formatting for APA style.
Research Context
A researcher tests whether a four-week mindfulness intervention reduces perceived stress compared to a waitlist control. Primary outcome: Perceived Stress Scale (PSS) score. Sample: 60 university students (30 per group). Hypothesized: Intervention > Control.
Step 1: Descriptive Statistics
First, describe your sample and baseline characteristics.
Participants’ demographic characteristics are presented in Table 1. The sample (N = 60) had a mean age of 21.3 years (SD = 2.1) and was 68% female. Baseline perceived stress scores did not differ between groups: intervention group M = 24.7 (SD = 4.3), control group M = 25.1 (SD = 4.1), t(58) = 0.31, p = .76.
Why this works: Reports sample size (N), means with SDs, compares baseline (shows randomization worked), includes statistical test with proper formatting.
Step 2: Inferential Statistics for Primary Outcome
Now the main hypothesis test.
The primary analysis tested whether the mindfulness intervention reduced perceived stress compared to control. As shown in Table 2, post-intervention stress scores were significantly lower in the intervention group (M = 17.2, SD = 3.8) than in the waitlist control group (M = 23.9, SD = 4.5). The independent-samples t-test confirmed this difference was statistically significant, t(58) = 5.21, p < .001, d = 0.86 (95% CI for d = 0.50-1.22). This represents a large effect size according to Cohen’s conventions.
Why this works:
- Restates hypothesis (“tested whether…”)
- Reports descriptive stats for both groups
- Gives full test results (t, df, p, d, CI)
- Interprets effect size magnitude (large)
- No speculation about why the effect occurred
Step 3: Secondary Outcomes (If Applicable)
Secondary outcomes included anxiety (GAD-7) and depression (PHQ-9). The intervention group showed significantly lower anxiety (M = 8.4, SD = 2.6) than controls (M = 12.1, SD = 3.0), t(58) = 4.37, p < .001, d = 0.72. Depression scores were also reduced, though the effect was smaller, t(58) = 2.89, p = .006, d = 0.45.
Step 4: Table That Accompanies This Text
The text above would be paired with a table like this (APA format):
Table 2
Mean Perceived Stress Scores by Treatment Group
| Group |
n |
M |
SD |
t |
p |
d |
95% CI for d |
| Intervention |
30 |
17.2 |
3.8 |
5.21 |
< .001 |
0.86 |
[0.50, 1.22] |
| Control |
30 |
23.9 |
4.5 |
|
|
|
|
Note: PSS = Perceived Stress Scale. CI = confidence interval.
Key points:
- Table title italicized, title case
- Abbreviations defined in note
- Values aligned properly
- Text references table and highlights key points without repeating every number
What NOT to Do (Bad Example)
The mindfulness program really helped people feel less stressed (p < .001), which shows that mindfulness works because it changes how people think about their problems. The participants in the treatment group had lower scores on the stress scale, which means they were less stressed. This is important because…
Problems:
- “really helped” – non-scientific language
- “p < .001” instead of exact p-value with test stat
- Interpretation (“shows that mindfulness works because…”) belongs in discussion
- Vague: no means, SDs, test statistic, effect size
- Subjective: no objective presentation
Checklist Before You Submit
Before finalizing your results section, run through this checklist:
Statistical Reporting
- All p-values reported exactly (p = .042, not p < .05)
- Effect sizes included for all key findings
- Test statistics, degrees of freedom, and p-values all present
- Confidence intervals reported for primary estimates
- N and n used correctly (total vs subgroup)
- Non-significant results included transparently
Formatting
- Statistics formatted correctly (italics for p, t, F, r, M, SD; no italics for ns)
- Decimal places consistent throughout
- Leading zeros used correctly (omit for p, d, r; use for t, F, M, SD)
- En dashes used for CI ranges (1.10-1.40), not hyphens or “to”
- Tables numbered consecutively (Table 1, Table 2, not Table One)
- Figures numbered consecutively (Figure 1, Figure 2)
- Every table and figure referenced in text, in order of appearance
- Table captions above tables; figure captions below figures
- Abbreviations defined on first use and in table/figure notes
Content and Organization
- Results section contains no interpretation or discussion of meaning
- Past tense used consistently for statistical findings
- Numbers in text match tables/figures exactly (no rounding discrepancies)
- Organized by research question/hypothesis, not by statistical test
- Descriptive statistics presented before inferential tests
- All relevant findings reported (no cherry-picking significant results)
- Style guide followed (APA/AMA/Chicago) throughout
Final Verification
- Checked all calculations: means, SDs, test statistics
- Verified that every statistical symbol is italicized correctly
- Ensured tables and figures are self-explanatory with complete captions
- Confirmed no raw data is presented in text (use tables for summaries)
- Verified that sample sizes match across text and tables
- Ensured no redundant repetition of table data in prose
Related Guides
For more on academic writing and research presentation, see these guides:
Conclusion
Effective data analysis writing for statistical results sections is not about clever prose or persuasive argument. It’s about clear, objective, precise reporting of what your data show. The results section is your evidence—treat it with the rigor it deserves. Follow your style guide’s formatting rules, report all relevant findings including non-significant ones, include effect sizes, and keep interpretation strictly out of results. When you separate what you found from what it means, you create research that is credible, replicable, and ready for publication.
Remember: Results = what happened. Discussion = why it matters. Master this distinction, and you’ll produce results sections that withstand peer review and contribute meaningfully to your field.
[^1]: Purdue OWL. (2023). Statistics in APA. https://owl.purdue.edu/owl/research_and_citation/apa_style/apa_formatting_and_style_guide/statistics_in_apa.html
[^2]: American Psychological Association. (2020). Publication Manual of the American Psychological Association (7th ed.). https://apastyle.apa.org/products/publication-manual-7th-edition
[^3]: University of Washington Writing Center. (n.d.). Reporting Results of Common Statistical Tests in APA Format. https://psych.uw.edu/storage/writing_center/stats.pdf
[^4]: Scribbr. (2020). Reporting Research Results in APA Style. https://www.scribbr.com/apa-style/results-section/
[^5]: APA Style. (2020). APA Style 7th Edition: Grammar Guidelines. https://apastyle.apa.org/instructional-aids/grammar-guidelines.pdf
[^6]: AMA Manual of Style. (2020). AMA Manual of Style: A Guide for Authors and Editors (11th ed.). https://www.amamanualofstyle.com/view/10.1093/jama/9780190246556.001.0001/med-9780190246556-chapter-19-div1-6
[^7]: NCBI/NIH. (2020). Guidelines for Statistical Reporting in Medical Journals. https://pmc.ncbi.nlm.nih.gov/articles/PMC7642026/
[^8]: Purdue OWL. (2023). Writing the Experimental Report. https://owl.purdue.edu/owl/subject_specific_writing/writing_in_the_social_sciences/writing_in_psychology_experimental_report_writing/experimental_reports_2.html
[^9]: Harhay, M. O., et al. (2020). Guidance on Statistical Reporting to Help Improve Your Research. https://pmc.ncbi.nlm.nih.gov/articles/PMC7193848/
[^10]: Verywell Mind. (2026). How to Write a Results Section. https://www.verywellmind.com/how-to-write-a-results-section-2795727
[^11]: APA Style. (2023). Numbers and Statistics Guide. https://apastyle.apa.org/instructional-aids/numbers-statistics-guide.pdf
[^12]: Purdue OWL. (2023). APA Numbers and Statistics. https://owl.purdue.edu/owl/research_and_citation/apa_style/apa_formatting_and_style_guide/apa_numbers_statistics.html
[^13]: APA Style. (2023). APA Tables and Figures. https://apastyle.apa.org/instructional-aids/tables-figures.pdf
[^14]: AMA Manual of Style. (2020). Study Design and Statistics. https://www.amamanualofstyle.com/view/10.1093/jama/9780190246556.001.0001/med-9780190246556-chapter-19-div1-6
[^15]: AMA Manual of Style. (2020). Common Statistical Abbreviations and Symbols. https://www.jcu.edu.sg/__data/assets/pdf_file/0003/1027227/Common-Statistical-Abbreviations-and-Symbols-in-AMA-7th-italics.pdf
[^16]: Taylor & Francis. (2023). American Medical Association (AMA) Style Guide. https://files.taylorandfrancis.com/tf_usama.pdf
[^17]: Chicago Manual of Style. (2017). Chicago-Style Citation Quick Guide: Author-Date. https://www.chicagomanualofstyle.org/tools_citationguide/citation-guide-2.html
[^18]: Chicago Manual of Style. (2017). The Chicago Manual of Style (17th ed.). https://www.chicagomanualofstyle.org
[^19]: Columbus State University. (2023). Reporting Quantitative Results in APA Style. https://www.columbusstate.edu/graduate-school/_docs/research-writing-bootcamp/fall-2023/reporting-quantitative-results-apa.pdf
[^20]: Harhay, M. O., et al. (2020). Guidance on Statistical Reporting to Help Improve Your Research. https://pmc.ncbi.nlm.nih.gov/articles/PMC7193848/
[^21]: APA Style. (2023). APA Tables and Figures. https://apastyle.apa.org/instructional-aids/tables-figures.pdf
[^22]: AMA Manual of Style. (2020). Tables and Figures. https://www.amamanualofstyle.com/view/10.1093/jama/9780190246556.001.0001/med-9780190246556-chapter-5-div1-2
[^23]: Purdue OWL. (2023). APA Tables and Figures. https://owl.purdue.edu/owl/research_and_citation/apa_style/apa_formatting_and_style_guide/apa_tables_and_figures.html
[^24]: Nature. (2024). Effective Writing: Figures and Tables. https://www.nature.com/scitable/effective-writing-figures-tables/
[^25]: Council of Science Editors. (2023). Style Manual. https://councilofscienceeditors.org
[^26]: APA Style. (2020). Publication Manual of the American Psychological Association (7th ed.). https://apastyle.apa.org/products/publication-manual-7th-edition
[^27]: AMA Manual of Style. (2020). Study Design and Statistics. https://www.amamanualofstyle.com/view/10.1093/jama/9780190246556.001.0001/med-9780190246556-chapter-19-div1-6
[^28]: Chicago Manual of Style. (2017). The Chicago Manual of Style (17th ed.). https://www.chicagomanualofstyle.org
[^29]: IEEE. (2023). IEEE Editorial Style Manual. https://ieeeauthorcenter.ieee.org
[^30]: APA Style. (2020). Bias-Free Language Guidelines. https://apastyle.apa.org/instructional-aids/bias-free-language-guidelines.pdf
[^31]: APA Style. (2020). Publication Manual of the American Psychological Association (7th ed.), pp. 118-120.
[^32]: Purdue OWL. (2023). APA Numbers and Statistics. https://owl.purdue.edu/owl/research_and_citation/apa_style/apa_formatting_and_style_guide/apa_numbers_statistics.html
[^33]: Based on guidelines from Scribbr, APA Style, AMA Manual, and NCBI statistical reporting recommendations.
Struggling with your results section? Get expert editing and detailed feedback from our academic writers who specialize in statistical reporting and style guide compliance. Visit QualityCustomEssays.com/essay-editing-service/ for professional assistance.
