APA 7th requires statistics reported with the symbol in italics, degrees of freedom in parentheses, the test value to two decimal places, an exact p value to two or three decimals, and an effect size. Report p < .001 only when the value is smaller than that; otherwise give the exact figure.
The General Rules
- Statistical symbols in italics: M, SD, t, F, r, p, N, n
- Greek letters are not italicised: α, β, χ²
- No leading zero where a value cannot exceed 1: p = .03, r = .42
- Leading zero where it can: M = 0.85
- Two decimals for most statistics; two or three for p
- Exact p values, except below .001, written as p < .001
- Report an effect size alongside every significance test
The most-corrected error: writing p = .000, which SPSS displays but which is not a real value. A probability is never zero. Report p < .001.
Independent-Samples t-Test
Participants who completed the training scored higher (M = 78.40, SD = 8.21) than those who did not (M = 71.60, SD = 9.04), t(98) = 3.94, p < .001, d = 0.79.
One-Way ANOVA
There was a significant effect of department on engagement, F(3, 196) = 5.62, p = .001, η² = .08.
Follow with the post hoc test and name it: "Tukey HSD comparisons indicated that…"
Correlation
Job satisfaction correlated positively with retention intention, r(148) = .42, p < .001.
The number in parentheses is degrees of freedom (N − 2), not the sample size.
Regression
The model was significant, F(3, 146) = 12.84, p < .001, and explained 21% of variance (R² = .21, adjusted R² = .19). Perceived support was the strongest predictor, β = .38, t(146) = 4.92, p < .001.
Report both R² and adjusted R²; with several predictors, adjusted is the honest figure.
Chi-Square
The association was significant, χ²(2, N = 200) = 9.44, p = .009, Cramér's V = .22.
Chi-square uniquely requires N inside the parentheses alongside degrees of freedom.
Effect Sizes and Why They Matter
- Cohen's d: 0.2 small, 0.5 medium, 0.8 large
- r: .10 small, .30 medium, .50 large
- η²: .01 small, .06 medium, .14 large
APA requires effect sizes because significance depends on sample size. With 5,000 participants a trivial difference reaches p < .001; the effect size is what tells your reader whether it matters.
Tables
Report in text where you have one or two results. Use a table for three or more, and never duplicate — if a value is in the table, do not repeat it in the sentence. APA tables use horizontal rules only, with a numbered title in italics above.
Frequently asked questions
Should I write p = .000 in APA?
No. A probability is never exactly zero, and p = .000 is an artefact of SPSS rounding the display. Report p < .001 instead. It is the single most frequently corrected statistical error in graduate papers.
Do I need a leading zero before a decimal in APA?
Only where the value can exceed 1. Write M = 0.85 with the zero, but p = .03 and r = .42 without it, because probabilities and correlations cannot exceed one.
Does APA require effect sizes?
Yes, alongside every significance test. Statistical significance depends on sample size, so with a large enough sample a trivial difference reaches p < .001. The effect size is what tells the reader whether the finding is practically meaningful.
What goes in the parentheses after t or F?
Degrees of freedom. A t-test takes one value, t(98); ANOVA and regression take two, F(3, 196), for between and within groups. Chi-square is the exception and also requires the sample size: χ²(2, N = 200).
Want This Checked Before You Submit?
Our tutors review your draft against your own rubric and explain what to change and why — used by 45,000+ students across the UK, USA, Canada and Australia.