Reporting Statistics in APA Style: Numbers, Tables and p-Values
How to report t-tests, ANOVA, correlations and regression in APA 7, exact formats, italicisation rules, effect sizes, and table construction.
Quantitative findings chapters lose marks less for wrong analysis than for wrong reporting. APA 7 prescribes exact formats for every common statistic, and because most disciplines borrow APA's conventions for numbers even under other referencing styles, these rules travel well beyond psychology.
General Number Rules
- Statistical symbols are italicised: M, SD, t, F, p, r, n, d. Greek letters (α, η²) are not.
- Numbers that can never exceed 1 (p-values, correlations, proportions) drop the leading zero: p = .03, r = .47. Everything else keeps it: d = 0.82.
- Report exact p-values to two or three decimals: p = .034. Use p < .001 only below that threshold, never "p = .000".
- Round descriptive statistics to two decimals. Percentages to whole numbers or one decimal.
Reporting the Common Tests
Descriptives: Participants in the intervention group scored higher (M = 24.3, SD = 4.1) than controls (M = 19.6, SD = 5.2).
Independent t-test: The difference was significant, t(58) = 3.86, p < .001, d = 0.99, 95% CI [2.28, 7.12].
One-way ANOVA: Teaching method affected exam scores, F(2, 87) = 6.24, p = .003, η² = .13. Tukey post-hoc comparisons showed…
Correlation: Study hours and grades were positively correlated, r(118) = .42, p < .001.
Chi-square: Completion differed by mode, χ²(1, N = 200) = 8.91, p = .003.
Regression: The model explained a significant proportion of variance, R² = .38, F(3, 96) = 19.4, p < .001. Sleep quality was the strongest predictor, β = .41, t = 4.52, p < .001.
The numbers in parentheses are degrees of freedom, omitting them is the most common student error.
Effect Sizes Are Not Optional
APA 7 requires effect sizes alongside significance. Conventional benchmarks: Cohen's d, 0.2 small, 0.5 medium, 0.8 large. Correlation r, .1/.3/.5. For η², .01/.06/.14. Interpret them in prose: statistical significance says an effect probably exists. The effect size says whether it matters.
Tables and Figures
- Number sequentially (Table 1, Figure 1). Label above the table with an italic title on the next line.
- Horizontal rules only, no vertical lines, no shading.
- Every table must be called out in text ("as shown in Table 2") and never merely duplicate what the prose already states: text gives the headline, the table gives the detail.
- Notes go below: Note. for general notes. Asterisks for probability notes (*p < .05. **p < .01.).
The Reporting Sequence for Any Result
- Restate the hypothesis or question.
- Name the test used.
- Give the statistic string (with df, p, effect size, CI).
- State the finding in plain English.
- Interpret magnitude, save theoretical meaning for the discussion chapter.
The Statistic Strings You Will Actually Need
Most of the reporting anxiety in a results chapter comes from not knowing the expected format. There are about eight strings, and once they are in front of you the chapter becomes mechanical.
| Test | Reported as |
|---|---|
| Independent t-test | t(58) = 2.41, p = .019, d = 0.63 |
| Paired t-test | t(39) = 3.12, p = .003, d = 0.49 |
| One-way ANOVA | F(2, 87) = 4.55, p = .013, η2 = .09 |
| Correlation | r(118) = .34, p < .001 |
| Chi-square | χ2(3, N = 240) = 9.87, p = .020, V = .20 |
| Linear regression | B = 0.42, SE = 0.11, β = .31, t = 3.82, p < .001 |
| Model fit | R2 = .28, F(3, 116) = 15.02, p < .001 |
| Odds ratio | OR = 2.40, 95% CI [1.60, 3.60], p = .002 |
| Descriptives | M = 4.21, SD = 0.88 |
Read the pattern rather than memorising the rows. Every string gives the test statistic, its degrees of freedom, the exact p-value and, where one exists, an effect size. Once you see that shape, an unfamiliar test is easy to format by analogy.
Italics, Case and Symbols
| Element | Formatting | Note |
|---|---|---|
| Statistical symbols in Latin letters | Italic | M, SD, t, F, p, r, N, d |
| Greek letters | Not italic | χ2, α, η2 |
| Vectors and matrices | Bold | Rare in student work |
| Total sample | Capital italic N | Subsample uses lower case n |
| Percentage | Use the symbol with a number | Write "percentage" when no number follows |
The Leading Zero Rule
This is the single most-missed convention in student results chapters, and it is entirely mechanical.
- No leading zero where the statistic cannot exceed 1: p-values, correlations, proportions, and alpha. Write p = .032, not 0.032.
- Keep the leading zero where the value can exceed 1: means, standard deviations, t and F values, and Cohen's d. Write M = 0.87.
Cohen's d catches people out, because it looks like an effect size and effect sizes feel bounded. It is not bounded, so it keeps its zero.
Reporting p-Values Correctly
Give exact values to two or three decimal places. The only case for an inequality is when the value falls below the smallest your output reports, where APA style uses p < .001.
Two habits to drop. First, "p = .000", which is what some software prints and which is never true, since a probability of exactly zero is not what the test computed. Write p < .001 instead. Second, describing a result as approaching significance. A p-value of .06 is not significant at the .05 level you set, and hedging it reads as reluctance to accept the result.
Also resist calling a result "highly significant" on the basis of a small p-value. The p-value indexes evidence against the null, not the size of the effect. That is what the effect size is for.
Tables and Figures
APA tables have a consistent anatomy: a number, a brief italic title, column headings, the body with no vertical rules, and notes beneath in the order general, specific, then probability. Horizontal rules appear above and below the headings and at the foot of the table, and nowhere else.
| Rule | Detail |
|---|---|
| Numbering | Table 1, Table 2, in order of first mention |
| Title | Italic, title case, brief but self-explanatory |
| Rules | Horizontal only. No vertical lines |
| Decimals | Consistent within a column |
| Reference in text | "as shown in Table 2", not "in the table above" |
| Duplication | Do not repeat every table value in the prose |
The last row matters most for readability. Text and table should divide the labour: the table carries the full numbers, the text points at what they show. Repeating the table in sentences doubles the length and halves the clarity.
Common Mistakes and Fixes
| Mistake | Fix |
|---|---|
| p = .000 | Write p < .001 |
| Leading zero on a p-value | Remove it. The value cannot exceed 1 |
| Degrees of freedom missing | Include them in brackets after the statistic |
| No effect size | Add one wherever the test has a standard companion |
| Interpreting theory in the results chapter | Report and describe here. Save meaning for the discussion |
| Copying SPSS output tables directly | Rebuild them in APA format. Raw output is not a results table |
| Inconsistent decimal places | Fix a convention, usually two, and hold it throughout |
What Belongs in Results, and What Does Not
The boundary between results and discussion is where a lot of otherwise good chapters lose marks. Results report what you found and describe its pattern and magnitude. The discussion explains why it might be so, how it relates to prior work, and what follows.
A quick test on any sentence in your results chapter: does it require a citation? If it does, it is interpretation and belongs in the discussion. Results sentences are supported by your own data.
Where Ethical Support Fits
Asking a statistics adviser to check that you have reported a model correctly, or having a results chapter proofread for consistency of formatting, is ordinary and expected. Formatting is presentation, not authorship.
What must stay yours is the analysis and the honest reporting of it. Round to the convention, never toward the threshold, and report every planned analysis, including the ones that returned nothing.
Frequently Asked Questions
How many decimal places should I use?
Two for most statistics, three for p-values. The underlying principle in APA style is to report to the precision the measurement supports, and to stay consistent.
Do I report non-significant results?
Yes, all of them, in the same format. Selective reporting distorts the record and is treated as a research integrity issue rather than a stylistic choice.
Should confidence intervals be included?
Where available, yes. APA style encourages them, and in health research they often carry more interpretive weight than the p-value.
Can I put statistics in the abstract?
Yes, and specific figures make an abstract far stronger. Keep the same formatting conventions there.
Where do assumption checks go?
Usually briefly at the start of the results, before the tests they support, or in the methods if you prefer. Either is acceptable, provided they appear somewhere.
My department uses Harvard, not APA. Do these rules apply?
The referencing style and the statistical reporting conventions are separable. Many departments use Harvard referencing while following APA conventions for statistics, because APA's are the most fully specified. Check your handbook.
Your Next Step Today
Search your results chapter for "0." immediately before a p-value or correlation, and for any instance of ".000". Those two searches catch the most common APA errors in under five minutes, and both are the kind of small mechanical fault that colours a marker's impression of the whole chapter.
Trusted Sources
- American Psychological Association, Numbers and Statistics Guide (PDF), the concise official reference. Accessed 11 August 2026.
- APA Style, Numbers. Accessed 11 August 2026.
- APA Style, Statistical abbreviations and symbols. Accessed 11 August 2026.
- APA Style, Tables and figures. Accessed 11 August 2026.
- University of Southern California, The Results, on the results and discussion boundary. Accessed 11 August 2026.
Always check the edition of the style manual your department requires. Where this guide and your handbook differ, follow your handbook.
Related services
Related guides
- APA 7th Edition Referencing: The Complete Guide
- Writing the Methodology Chapter: Design, Sampling and Justification
- Dissertation Structure: A Chapter-by-Chapter Blueprint
- Medical Research: Data Collection, Analysis and Reporting
All academic writing guides or browse our academic writing services.