How to Report Statistics in APA Style (7th ed.): Examples for Each Test
Reviewed by Prof. Dr. Alexander Holtermann · Updated 16 October 2026 · 6 min read
APA style reports test statistic, degrees of freedom, exact p-value and effect size, e.g. t(13.46) = 3.42, p = .004, d = 1.71. Examples for every test.
Key takeaways
- A complete result contains descriptive statistics, the test statistic with degrees of freedom, the exact p-value, an effect size and ideally a confidence interval.
- Statistical symbols such as M, SD, t, F, p, r and d are italicized; Greek letters (α, β, χ², η²) are not.
- Use no leading zero for values that cannot exceed 1 (p = .004, r = .84, α = .91) and a leading zero otherwise (d = 0.63, M = 0.85).
- Report exact p-values with two or three decimals; for very small values write p < .001, never p = .000.
- Round to two decimals in most cases; use a table instead of the text when you have more than about three numbers per result.
The basic pattern of an APA result
APA style (Publication Manual, 7th edition, 2020) aims at results that are complete, compact and comparable. Every inferential result follows the same pattern: a sentence in plain language stating what you found and in which direction, followed by the statistics.
statistic(df) = value, p = value, effect size = value, 95% CI [lower, upper]
The sentence comes first, the numbers support it. Readers should understand the finding even if they skip the parentheses. The statistics must allow them to check your conclusion and to use your result in a meta-analysis.
- Descriptive statistics for each group or variable (M and SD, or Mdn and IQR, or n and %).
- Test statistic with its degrees of freedom: t(14), F(2, 12), χ²(1, N = 100), r(8).
- Exact p-value: p = .004 rather than p < .05.
- Effect size: d, r, η², φ, R² or an odds ratio.
- Confidence interval for the main estimate or effect size, where available.
Formatting rules for numbers and symbols
These rules from the APA manual cause most of the formal corrections in theses.
| Rule | Correct | Wrong |
|---|---|---|
| Statistical symbols are italic (Latin letters) | M, SD, t, F, p, r, n, d (in italics) | M, SD in plain font |
| Greek letters and abbreviations are not italic | α, β, χ², η², CI | italic CI |
| No leading zero if the value cannot exceed 1 | p = .032, r = .45, α = .84 | p = 0.032 |
| Leading zero if it can exceed 1 | d = 0.63, M = 0.85, t = 0.94 | d = .63 |
| Exact p-values, two or three decimals | p = .004, p = .27 | p = .000, p < .05 for every result |
| Very small p-values | p < .001 | p = .0000034 |
| Spaces around = and < | t(14) = 3.42 | t(14)=3.42 |
| Decimal point, not comma | M = 16.50 | M = 16,50 |
| Brackets for CIs | 95% CI [0.92, 4.08] | 95% CI (0.92–4.08) |
| N for the total sample, n for subgroups | N = 100; n = 50 per group | n = 100 total |
Examples for t-tests
Always report the means and SDs of both groups or both time points. For Welch’s t-test the df have decimals.
- Independent (Welch): “Students with a study plan (M = 16.50, SD = 1.60) scored higher than students without one (M = 14.00, SD = 1.31), t(13.46) = 3.42, p = .004, d = 1.71, 95% CI [0.92, 4.08].”
- Paired: “Exam anxiety decreased from before (M = 24.50, SD = 3.83) to after the training (M = 20.67, SD = 3.27), t(5) = 5.86, p = .002, d_z = 2.39.”
- One sample: “Students studied more than the planned 20 hours per week (M = 23.00, SD = 2.58), t(9) = 3.67, p = .005, d = 1.16.”
- Non-significant: “Satisfaction did not differ significantly between groups, t(58) = 0.94, p = .35, d = 0.24, 95% CI [−0.27, 0.75].” Report the effect size even here.
Examples for ANOVA, chi-square and non-parametric tests
The structure stays the same; only the statistic changes.
- One-way ANOVA: “Teaching format had a significant effect on quiz scores, F(2, 12) = 9.50, p = .003, η² = .61. Tukey post-hoc tests showed that the flipped classroom group (M = 16.00, SD = 1.58) scored higher than the self-study group (M = 11.00, SD = 2.24), p = .003.”
- Chi-square: “Passing was associated with study group membership, χ²(1, N = 100) = 9.89, p = .002, φ = .31; 80% of study group members passed compared with 50% of non-members.”
- Mann-Whitney U: “In-person exam students reported more stress (Mdn = 7) than online exam students (Mdn = 4), U = 3.00, z = −2.78, p = .005, r = .74.”
- Goodness of fit: “The three exam dates were not chosen equally often, χ²(2, N = 60) = 8.40, p = .015.”
Examples for correlation, regression and reliability
Correlations have df = n − 2. For regression, report the model test and the coefficients.
- Pearson: “Study hours were positively correlated with exam scores, r(8) = .84, p = .002, 95% CI [.44, .96].”
- Spearman: “r_s(8) = .79, p = .006.”
- Simple linear regression: “Study hours significantly predicted exam scores, F(1, 8) = 18.80, p = .002, R² = .70; b = 1.60, 95% CI [0.75, 2.44].”
- Logistic regression: “Attending the review session was associated with higher odds of passing, OR = 3.00, 95% CI [1.38, 6.50], p = .005.”
- Cronbach’s alpha: “The four-item exam anxiety scale showed excellent internal consistency, α = .91.”
Tables and figures
When a paragraph would contain more than a handful of numbers, use a table. APA tables have a number in bold (Table 1), a title in italic title case on the next line, horizontal lines only (no vertical lines), and notes below the table explaining abbreviations and significance markers.
| Variable | n | M | SD | 1 | 2 |
|---|---|---|---|---|---|
| 1. Study hours per week | 10 | 7.80 | 3.94 | — | |
| 2. Exam score | 10 | 65.90 | 7.50 | .84** | — |
Checklist before you submit
Go through your results section with this list. It catches most formal errors.
- Every hypothesis has a result sentence, including non-significant ones.
- Every test has df, exact p and an effect size.
- Descriptive statistics are given for all groups compared.
- Leading zeros and decimal places are consistent.
- Statistical symbols are italic, Greek letters are not.
- Assumption checks and any corrections (Welch, Bonferroni, transformations) are mentioned.
- Interpretation (what it means) is in the discussion, not in the results section.
Frequently asked questions
How do I report a p-value in APA 7?
Exactly, with two or three decimals and no leading zero: p = .032. For values below .001 write p < .001. The p is italicized.
Do I need to report effect sizes in APA style?
Yes. APA 7 recommends reporting an effect size, ideally with a confidence interval, for every main result, whether significant or not.
How many decimal places should I use?
Usually two for means, SDs, test statistics and effect sizes. p-values with two or three decimals. Percentages as whole numbers or with one decimal.
Why no leading zero for p and r?
APA omits the zero before the decimal point for statistics that cannot be greater than 1, such as p, r, α and η². Values that can exceed 1, like d or M, keep it.
Can I write p < .05 instead of the exact value?
Only in tables with significance markers. In the text, APA 7 asks for exact p-values, except below .001.
Sources and further reading
- American Psychological Association. (2020). Publication Manual of the American Psychological Association (7th ed.).
- Wasserstein, R. L., & Lazar, N. A. (2016). The ASA statement on p-values: Context, process, and purpose. The American Statistician, 70(2), 129–133.
Your university’s examination regulations and style guide always take precedence.
Related articles
- APA 7th editionAPA 7 uses author–date citations: (Nguyen & Patel, 2021) or Nguyen and Patel (2021). Use “et al.” for three or more authors from the first citation.
- Tables and figuresIn APA 7, label tables and figures in bold (Table 1), add an italic title in title case above, put notes below, and cite any source in a copyright note.
- p-value and significanceThe p-value is the probability of a result at least as extreme as yours if the null hypothesis were true. p < .05 is usually called significant.
- Effect sizeAn effect size tells you how large a difference or relationship is, independent of sample size. Cohen’s d of 0.2, 0.5 and 0.8 is small, medium, large.
- Confidence intervalA 95% confidence interval gives a range of plausible values for a population parameter. It shows both the size of an estimate and its precision.
- t-test in SPSSIn SPSS: Analyze → Compare Means → Independent-Samples T Test. Read Levene’s test, then t, df, p and the CI, and report with Cohen’s d in APA style.
