How to Report Statistics in APA Format (With Examples, 2026)
Reporting statistics in APA format comes down to three rules. Put a zero before the decimal only when the statistic can exceed 1. Report most statistics to two decimal places, and give exact p values to two or three decimals unless they fall below .001. Everything else is applying those consistently.
The rules are small and the penalty for getting them wrong is disproportionate. A manuscript with p = 0.03 instead of p = .03 signals that nobody checked the details. The reviewer forms that impression before reading a single result.
This guide covers the decimal and symbol rules from APA's own guidance, and how to lay out a results table. It ends with where doing it by hand starts costing more than it should.
What you need before you start
You need your analysis output: means, standard deviations, test statistics, degrees of freedom and p values. You also need to know which test produced each number. APA formatting is not a find-and-replace pass; a t statistic and a correlation follow different decimal rules.
You also need your target journal's guide. APA Style is the baseline for psychology, education and much of the social sciences, but journals layer their own requirements on top. Check before you format everything twice.
The reference used throughout here is APA's own Numbers and Statistics Guide for the 7th edition. It condenses the relevant Publication Manual sections into a few pages. It is worth keeping open.
How to format the numbers correctly
Option 1: Get the decimals right
The leading-zero rule is the one reviewers notice first. APA states it directly: put a zero before the decimal point when a number is less than 1 but the statistic can exceed 1. Do not use a zero before a decimal when the statistic cannot be greater than 1 — proportions, correlations, and levels of statistical significance.
So a mean of 0.82 keeps its zero, because means can exceed 1. A p value of .03 does not, and neither does a correlation of .45 or an R² of .12.
For decimal places, APA gives three defaults. Report means and standard deviations for data measured on integer scales, such as surveys and questionnaires, to one decimal. Report other means, standard deviations, correlations, proportions and inferential statistics — t, F, chi-square — to two decimals. Report exact p values to two or three decimals, for example p = .006 or p = .03.
Below .001, stop reporting exact values. APA's instruction is to report p values less than .001 as p < .001. The overarching principle it gives is to round as much as possible while considering prospective use and statistical precision.
Option 2: Format the symbols
Statistical symbols used as letters are italicised: M, SD, t, F, p, r, N, n, df. Greek letters are not — APA specifies standard, non-italic type for those.
Two rules about wording are easy to miss. Use the symbol or abbreviation when there is a mathematical operator attached, as in M = 7.7. Use the term rather than the symbol in running text, so you write "the means were similar" and not "the Ms were similar."
You also do not need to define standard statistical abbreviations. APA lists M, SD, F, t, df, p, N, n and OR among those that need no definition. Other abbreviations — ANOVA, CI, RMSEA, SEM — do get defined on first use.
One typographic detail: for a negative value, put a space only before the minus sign, not after it.
Option 3: Build the results table
A results table exists so readers do not have to reconstruct numbers from prose. Put the descriptive statistics there — group, n, mean, standard deviation — and keep the inferential result in the text where you interpret it.
Assembled correctly, a comparison reads as t(48) = 2.31, p = .025, d = 0.62. Degrees of freedom go in parentheses immediately after the symbol, and the test statistic takes two decimals. The p value takes no leading zero. The effect size follows, because significance alone does not tell a reader whether the difference matters.
Keep the decimal places consistent down each column. Mixing one and two decimals in the same column is the most common formatting error in submitted tables.
One more detail on numerals. APA uses figures for numbers 10 and above and for all numbers used in statistics. Figures also cover numbers attached to units of measurement, scores, ages, dates and table or figure numbers. Numbers below 10 are spelled out, as are numbers that begin a sentence. So a sentence opens with "Sixty participants" and a result reads "2.45," and both are correct.
Where the manual route slows down
The formatting itself is quick. The transcription is not, and that is where errors enter.
Numbers get copied from statistical output into a table, then quoted again in the text, then again in an abstract. Every copy is a chance for a digit to drift, and by the third copy nobody remembers which version is authoritative. Reviewers find these inconsistencies reliably, because comparing a table against its own prose is the easiest check they can run.
Revision makes it worse. Drop two outliers and rerun. Every mean, standard deviation, test statistic and p value in three places then has to be updated by hand. Miss one and the manuscript contradicts itself.
How to format APA statistics with Powerdrill Bloom
Step 1: Upload your data
Upload the dataset or your analysis output. Powerdrill Bloom profiles the columns, so group sizes and missing values are visible before any statistic is computed. The numbers that end up in your table start from a base you have actually seen.
Step 2: Describe what you need in natural language
Ask for the comparison and the format together. Compare these two groups, report the t test with degrees of freedom, exact p value and effect size, formatted to APA 7th edition conventions. Ask for the descriptive table by group in the same pass, so the table and the sentence come from one calculation rather than two transcriptions.
Step 3: Export the chart, report, or deck
Take out the formatted table, the accompanying figure, or a written results paragraph you can paste into the manuscript.
Why this beats reformatting after every revision
| Manual route | Powerdrill Bloom | |
|---|---|---|
| Rerunning after excluding cases | Update every number by hand in three places | Re-ask on the filtered data |
| Table and text agreement | Manual cross-check | Both from one calculation |
| Decimal consistency | Per-cell discipline | Stated once in the request |
| Effect size alongside p | Separate calculation | Ask for it in the same pass |
The first row is where the time goes. Manuscripts get revised several times before acceptance, and each revision means the whole reporting layer is rebuilt. Making that cheap is what keeps the table and the text honest with each other.
Common mistakes
Writing p = 0.03. The leading zero is wrong because p cannot exceed 1. The same applies to correlations, proportions and R².
Reporting p = .000. Statistical software prints this when rounding, but APA requires p < .001. A p value is never exactly zero, so the printed form is an artefact.
Mixing decimal places within a column. Pick the rule for that statistic and apply it to every row.
Reporting significance without an effect size. A p value says the difference is probably not zero, not that it is large enough to matter — see what statistical significance is.
Italicising Greek letters. Latin statistical symbols are italic; Greek letters stay upright. This one is easy to introduce accidentally when pasting from software output.
Conclusion
APA statistical formatting rewards consistency more than memorisation. Leading zeros only where the statistic can exceed 1. Two decimals for most inferential statistics, exact p values down to .001 and p < .001 below that, italics for Latin symbols but not Greek.
The part worth automating is not the formatting but the rebuild after every revision. Try Powerdrill Bloom on your dataset so the table and the sentence come from the same calculation. See also our academic papers tool, the guide to making charts for a research paper, and running descriptive statistics.
Frequently asked questions
How do I report a p value in APA format?
Give the exact value to two or three decimals with no leading zero, as in p = .006 or p = .03. For values below .001, report p < .001 rather than an exact figure. Never write p = .000.
Does APA use a zero before the decimal point?
Only when the statistic can exceed 1. Means and standard deviations keep the zero; p values, correlations, proportions and R² do not, because those cannot be greater than 1.
How many decimal places should I use?
One decimal for means and standard deviations from integer scales such as questionnaires. Two decimals for other means and standard deviations, correlations, proportions and inferential statistics including t, F and chi-square.
Which statistical symbols are italicised in APA?
Latin letters used as statistical symbols — M, SD, t, F, p, r, N, n, df — are italicised. Greek letters are set in standard non-italic type.
Do I need to define abbreviations like M and SD?
No. APA treats standard statistical symbols such as M, SD, F, t, df, p, N, n and OR as needing no definition. Other abbreviations, including ANOVA, CI and RMSEA, are defined on first use.