How to report a t-test in APA style
APA style gives the test statistic, its degrees of freedom, the exact p-value, an effect size and a confidence interval, in a fixed order. Below are the rules and example sentences generated from real results.
Sources checked on November 21, 2026. Results computed by the site's engine and tested against R and SciPy.
The pattern
t(df) = value, p = value, d = value, 95% CI [lower, upper]
The APA Numbers and Statistics Guide sets these conventions:
- Report t and effect sizes to two decimals.
- Report exact p-values to two or three decimals; we use three. Below .001, write p < .001. Never write p = .000.
- Do not put a zero before the decimal point for statistics that cannot exceed 1, such as p and correlations. Do put it for t and d.
- Italicize statistical symbols: t, p, d, M, SD, n.
- Give Welch's degrees of freedom to two decimals; integers stay whole.
- Use a minus sign for negative values.
Independent samples (Welch)
Example from R's mtcars data:
Cars with manual gearboxes had higher fuel economy (M = 24.39, SD = 6.17) than cars with automatic gearboxes (M = 17.15, SD = 3.83), Welch's t(18.33) = −3.77, p = .001, g = −1.44, 95% CI [−11.28, −3.21].
Name the test as Welch's, because readers otherwise assume Student's. The confidence interval is for the difference in means, in the original units (miles per gallon here).
Paired samples
Patients gained less extra sleep with drug 1 (M = 0.75, SD = 1.79) than with drug 2 (M = 2.33, SD = 2.00), t(9) = −4.06, p = .003, dz = −1.28, 95% CI [−2.46, −0.70].
Say which effect size you used. For paired data dz and dav can differ a lot (here dav = −0.83).
One-sample
The mean extra sleep across all observations (M = 1.54, SD = 2.02) was greater than zero, t(19) = 3.41, p = .003, d = 0.76, 95% CI [0.60, 2.48].
Non-significant results
Treated as independent groups, the sleep data showed no significant difference between the drugs, Welch's t(17.78) = −1.86, p = .079, g = −0.80, 95% CI [−3.37, 0.21].
Do not write "there was no difference". Report the interval, which here includes differences from about −3.4 to 0.2 hours, so a sizeable effect is still possible.
One-sided tests
State the direction and that the test was one-sided, for example: "t(9) = −4.06, p = .001, one-tailed". Justify the one-sided test in the methods section. See one-tailed vs two-tailed tests.
Common slips to check before you submit
- Writing p = .000 when software prints 0.000. Software rounds; the true p is never zero. Write p < .001.
- Rounding Welch's df to a whole number in the report while the p came from the decimal df. Report the df you used, here 18.33.
- Giving the CI of each group mean instead of the CI of the difference. Readers need the difference: −11.28 to −3.21 mpg in the gearbox example.
- Dropping the sign of t. A negative t only says which group was listed first; keep it, and describe the direction in words.
- Calling a one-tailed p two-tailed, or halving a p after seeing the data. State the alternative you planned.
Tables and many tests
When you report several comparisons, a table with columns for each group's M and SD, then t, df, p, the effect size and the CI keeps the text readable. If you run many tests, say how you adjusted for multiple comparisons. Every result on this site includes an APA line you can copy and adapt; check it against your own journal's style, since some ask for two-decimal p-values.
Sources
- APA Style, Numbers and Statistics Guide (7th edition handout) checked November 21, 2026