The p value is the most-reported number in quantitative research and the one most often formatted wrongly. APA 7's rules are short. This guide lists them with examples, and covers the wording around significant and non-significant results.
The APA 7 format template
[Test result], t(df) = X.XX, p = .XXX, [effect size].
All 53 templates, ready to paste with the italics set: the Rescue Vault, $27 →
The p value always follows a test statistic and its degrees of freedom. On its own, "(p = .03)" tells the reader too little.
The rules
- Report exact values: p = .032, not p < .05.
- Two or three decimals: p = .03 or p = .032. Be consistent in the paper; three is more common.
- Below .001, write p < .001. Never p = .000: software prints .000 when it rounds, and a p value is never zero.
- No leading zero: p = .032, not 0.032, because p cannot exceed 1.
- Italicize p (it is a statistical symbol); spaces around = and <.
- Very large values: p = .97 is fine; software showing 1.000 after a correction should be reported as p > .999.
Worked examples
Significant result:
Students who used spaced practice recalled more words than students who used massed practice, t(22) = 2.32, p = .030, d = 0.95.
Very small p value:
Adherence was strongly correlated with weight loss, r(73) = .64, p < .001.
Non-significant result:
The low-carb and Mediterranean diets did not differ significantly, p = .964.
How to word results around the p value
- Significant: "a significant difference", "significantly higher", with the direction.
- Non-significant: "no significant difference was found". Not "there was no difference" (absence of evidence is not evidence of absence), and not "a trend towards significance" or "marginally significant" for p = .06.
- Report the effect size next to p: a tiny p in a large sample can come with a trivial effect.
Mistakes reviewers catch
- p = .000 instead of p < .001.
- p < .05 when the exact value is known. Thresholds belong in table notes.
- A leading zero: 0.032.
- p values that do not match their statistic. About half of psychology papers contain at least one. See how often p values are misreported.
- Unadjusted p values for many comparisons. Report the corrected ones. See post hoc tests.
Let KyroStat do the write-up for you
KyroStat formats every p value by these rules and computes each one from the test statistic, so the numbers in your Results always agree with each other.
Frequently asked questions
Two or three decimals? APA 7 allows either; three is more common in psychology. Use the same number throughout.
One-tailed or two-tailed? Two-tailed is the default. If you planned a one-tailed test, say so and why.
What about asterisks in tables? In tables you can mark significance with asterisks and explain them in the note (*p < .05. **p < .01. ***p < .001.). See how to format APA tables.