How to Report Statistical Results (With APA Examples)
The standard formula for reporting a statistical result is: test statistic + degrees of freedom + p-value + effect size, followed by one plain-language sentence saying what the finding means. For example: “The experimental group scored significantly higher than the control group, t(58) = 2.31, p = .024, d = 0.61.” Writing only “p < .05, significant” is an incomplete report — committees and reviewers expect the exact p-value and an effect size.
This guide collects APA-style write-ups for the most common tests, shows how to structure a results section, and lists the reporting mistakes reviewers catch first.
Who is this guide for?
Thesis students whose analyses are done and who are now writing the results chapter
Manuscript authors who received a “statistical reporting is insufficient” reviewer note
Anyone unsure how to translate SPSS output into academic prose
Researchers preparing APA 7–compliant tables and text
General rules that apply to every test
Whatever test you ran, these rules set the quality of your report:
Statistical symbols are italicized: t, F, p, r, d, M, SD. (Greek letters are not: χ², η².)
Report exact p-values: p = .024. Reserve “p < .05” for tables and grouped displays.
Never copy SPSS's “p = .000” — the correct form is p < .001.
No leading zero for statistics that cannot exceed 1 (p, r, η²): p = .024, r = .42. Keep it for those that can: d = 0.61, t = 2.31.
Report an effect size with every main finding: Cohen's d, η²p, r, Cramér's V — whichever fits your test.
Give descriptives before the inference: group means and standard deviations (M, SD) let the reader make sense of the result.
Never leave a number alone — every statistical sentence gets a plain-language companion stating what the finding means.
APA write-ups, test by test
Copy-ready example sentences for the most common tests — swap in your own numbers:
| Test | Example APA write-up |
|---|---|
| Independent-samples t-test | The experimental group (M = 78.4, SD = 8.2) scored significantly higher than the control group (M = 71.9, SD = 9.1), t(58) = 2.31, p = .024, d = 0.61. |
| Paired-samples t-test | Post-test scores were significantly higher than pre-test scores, t(29) = 4.12, p < .001, d = 0.75. |
| One-way ANOVA | A significant difference was found between groups, F(2, 87) = 5.42, p = .006, η²p = .11. Tukey's test located the difference between groups A and C (p = .004). |
| Chi-square test of independence | A significant association was found between gender and preference, χ²(1, N = 120) = 6.85, p = .009, Cramér's V = .24. |
| Pearson correlation | Anxiety was significantly and negatively related to academic achievement, r(98) = -.42, p < .001. |
| Multiple linear regression | The model was significant, F(3, 116) = 14.20, p < .001, explaining 27% of the variance (R² = .27). The strongest predictor was self-efficacy, β = .38, t = 3.20, p = .002. |
| Mann-Whitney U | Groups differed significantly, U = 245.5, z = -2.10, p = .036, r = .27. |
| Cronbach's alpha | The scale showed high internal consistency (α = .87). |
How to structure the results section
A good results section carries the reader from data to conclusion in stages:
- 01
Introduce the sample
Start with participant counts, demographic breakdowns and descriptive statistics for the scales (M, SD, range) — usually in a single table.
- 02
Report assumption checks
Briefly state the outcome of normality (Shapiro-Wilk, skewness-kurtosis) and homogeneity (Levene's) checks, and which test you therefore chose.
- 03
Answer each research question in order
Follow the order of your research questions: restate the question, report the relevant test in APA style, and say in one sentence what the finding means.
- 04
Support with tables and figures
Move comparisons with more than three numbers into a table; refer to it in text (“As shown in Table 3...”). Don't repeat the same values in both table and text.
The most common reporting mistakes
These are the errors committees and reviewers catch first:
Writing p = .000 — it's SPSS rounding; the correct form is p < .001.
Omitting effect sizes — “significant” doesn't mean the effect is large; a tiny effect turns significant in a big enough sample.
Conflating “significant” with “important” — statistical significance is not a claim of practical importance; keep the distinction in your interpretation.
Pasting SPSS tables straight into the thesis — SPSS output is not APA table format; tables must be rebuilt.
Reporting only significant results — non-significant findings are reported too; an unsupported hypothesis is still a finding.
Inconsistent decimals — statistics usually take two decimals (t = 2.31), p-values two or three (p = .024); keep one scheme throughout.
Causal language — interpreting correlational or cross-sectional findings as “affected, caused”; say “was associated with” instead.
Frequently asked questions
SPSS shows my p-value as .000 — how should I write it?
Write p < .001. SPSS rounds to three decimal places, which produces .000; a p-value can never be exactly zero, so .000 is considered an error.
Should I report non-significant results?
Yes, always. An unsupported hypothesis is still a finding and is reported in the same format: t(58) = 1.12, p = .268, d = 0.29. Reporting only significant results counts as selective reporting and is an ethical problem.
Is reporting effect sizes mandatory?
APA 7 and most journal guidelines explicitly require them, and many committees ask. Cohen's rough benchmarks: for d, 0.2 is small, 0.5 medium, 0.8 large; for η²p, .01 small, .06 medium, .14 large. Which measure to report depends on the test used.
Can I put SPSS tables directly into my thesis?
No. SPSS output tables are not APA-formatted: they carry unnecessary columns and the wrong layout. Tables should be rebuilt to APA 7 conventions (italic title above, minimal horizontal rules, only the needed columns). Tables in GetBayes reports are delivered ready to drop into a thesis.
How many decimal places should I use?
The common convention: means, standard deviations and test statistics take two decimals (M = 78.42, t = 2.31); p-values two or three (p = .024); correlations two (r = .42). Whatever scheme you choose, apply it consistently throughout.
Can you check my results section for me?
Yes. If we run your analyses, the report is delivered in this format already; if you ran your own, we can review and correct the reporting. Send your data and current text — we'll reply with a free assessment.
Make your results section publication-ready
Our analysis reports come with APA-formatted tables and academically written interpretation — ready for your defense and your journal. Send your data and we'll reply within 24 hours.
Last updated: July 8, 2026