How to Report a t-Test in APA 7 Without Making It Weird
Reporting a t test in APA 7 is not difficult, which is exactly why it becomes annoying. This guide shows you what to include, what to leave out, and how to write independent, paired-samples, and one-sample t tests without producing the usual statistics goblin sentence.
Free article tool included
There’s a free APA t-test reporter near the end of this guide. Read the examples first, or jump straight to the tool if you already have your means, standard deviations, t value, degrees of freedom, and p value.
A t test write-up should be boring in the best possible way. It should tell the reader what you compared, what the means were, whether the difference was statistically significant, and what the test result was. That is the job. No suspense. No interpretive fog. No “this proves that humans are complex creatures” paragraph creeping into the Results section like it owns the place.
The problem is that statistical output rarely arrives in a form you can paste straight into your assignment. SPSS, JASP, jamovi, Excel, and R all give you numbers, but they do not always give you a clean APA sentence. So students often end up with something that is either too thin:
“The t test was significant.”
Or too much:
“The independent-samples t test showed that the mean anxiety score for the caffeine group was statistically significantly higher than the mean anxiety score for the no-caffeine group, suggesting that caffeine may have had a meaningful psychological effect on anxious arousal in participants.”
That second one is already drifting into Discussion-section territory with its little suitcase packed.
This guide keeps it simpler. Here is what you need, where it goes, and how to stop the sentence becoming weird.
The basic APA 7 format for a t test
A standard APA-style t test report usually includes:
The type of comparison or result in plain English.
The descriptive statistics, usually the mean and standard deviation for each group or condition.
The t value.
The degrees of freedom in parentheses.
The p value.
An effect size, often Cohen’s d, if required by your course, supervisor, journal, or marking rubric.
The basic structure looks like this:
There was a significant difference between Group A and Group B, with Group A scoring higher than Group B, t(df) = value, p = value, d = value.
A cleaner version with descriptives looks like this:
Participants in the caffeine condition reported higher anxiety scores (M = 6.42, SD = 1.31) than participants in the no-caffeine condition (M = 4.98, SD = 1.26), t(58) = 4.31, p < .001, d = 1.13.
That is the general shape. The sentence can change depending on the type of t test, but the bones are the same.
A small formatting note before we go further: in APA style, statistical symbols such as t, p, M, SD, and d are italicised. The title says “t-Test” because search engines are practical creatures, not elegant ones. In the write-up itself, use APA-style statistical formatting.
Independent-samples t testexample
Use an independent-samples t test when you are comparing two separate groups. For example, psychology students versus sociology students, caffeine group versus control group, or therapy group versus waitlist group.
Imagine you compared stress scores between students who worked part-time and students who did not.
The output gives you:
Part-time students: M = 22.40, SD = 5.18
Non-working students: M = 19.65, SD = 4.92
t(78) = 2.43
p = .017
Cohen’s d = 0.54
A good APA write-up would be:
Students who worked part-time reported higher stress scores (M = 22.40, SD = 5.18) than students who did not work part-time (M = 19.65, SD = 4.92), t(78) = 2.43, p = .017, d = 0.54.
That sentence does four useful things. It names the groups, gives the direction of the difference, reports the descriptive statistics, and includes the inferential test result. Lovely. Slightly dull. Exactly the point.
Paired-samples t testexample
Use a paired-samples t test when the same participants are measured twice, or when scores are naturally paired. For example, anxiety before and after an intervention, reaction time in condition A and condition B, or memory scores before and after sleep deprivation has done its little damage.
Imagine you measured confidence before and after a short presentation-skills workshop.
Before workshop: M = 4.10, SD = 1.22
After workshop: M = 5.35, SD = 1.08
t(39) = 5.62
p < .001
Cohen’s d = 0.89
A clean APA write-up would be:
Participants reported higher confidence after the workshop (M = 5.35, SD = 1.08) than before the workshop (M = 4.10, SD = 1.22), t(39) = 5.62, p < .001, d = 0.89.
Do not just write “there was a significant difference between before and after.” That is technically a result, but it is not very helpful. Say which score was higher. Your reader should not have to go hunting through the table like this is an academic escape room.
One-sample t test example
Use a one-sample t test when you are comparing one sample mean against a known value, benchmark, scale midpoint, population value, or theoretical value.
Imagine students completed a wellbeing scale where the scale midpoint was 20. You want to know whether the sample scored significantly above that midpoint.
Sample: M = 23.70, SD = 6.15
Comparison value: 20
t(49) = 4.25
p < .001
Cohen’s d = 0.60
A good APA write-up would be:
Participants’ wellbeing scores (M = 23.70, SD = 6.15) were significantly higher than the scale midpoint of 20, t(49) = 4.25, p < .001, d = 0.60.
Again, the sentence does not need to perform. It needs to report.
What if the t test is not significant?
A non-significant result is still a result. It is not a failed statistic. It is not the universe personally rejecting your dissertation. It simply means the test did not provide sufficient evidence of a statistically significant difference.
Suppose your independent-samples t test gives you:
Group A: M = 18.40, SD = 4.02
Group B: M = 17.95, SD = 4.31
t(62) = 0.43
p = .671
Cohen’s d = 0.11
You could write:
There was no significant difference in memory scores between Group A (M = 18.40, SD = 4.02) and Group B (M = 17.95, SD = 4.31), t(62) = 0.43, p = .671, d = 0.11.
Keep the same core information. You still report the means. You still report the test statistic. You still report the p value. You do not bury the result in shame.
What you should avoid is this:
The hypothesis was wrong.
That is too blunt and not really how hypothesis testing works.
Also avoid this:
There was no difference between the groups.
That is stronger than your test can usually justify. A safer phrasing is:
There was no significant difference between the groups.
That wording says what the test found without pretending your sample has personally audited reality.
Should you report Cohen’s d?
Usually, yes, especially in psychology. A p value tells you whether the result is statistically significant under the test model. It does not tell you how large or useful the difference is. Cohen’s d gives a standardised estimate of the difference between means.
For a t test, the result might be statistically significant but small, especially with a large sample. It might also be non-significant but still show a potentially interesting effect in a small, underpowered study. This is why many psychology courses expect an effect size alongside the test result.
A complete write-up might look like this:
Participants in the intervention group reported lower anxiety scores (M = 12.10, SD = 3.42) than participants in the control group (M = 14.85, SD = 3.88), t(70) = -3.16, p = .002, d = 0.75.
Notice that the t value is negative here. That is not a problem. The sign depends on the order in which the groups were entered or compared. The sentence gives the meaningful direction of the effect, so the reader does not have to decode the minus sign like it is a cursed rune.
How many decimal places should you use?
For most student APA write-ups, a sensible pattern is:
Means and standard deviations: two decimal places.
t values: two decimal places.
p values: three decimal places when exact, unless p < .001.
Effect sizes: usually two decimal places.
So you would write:
Good:t(48) = 2.36, p = .022, d = 0.68
Good:t(48) = 4.91, p < .001, d = 1.42
Less good:t(48) = 2.363829, p = 0.0217359
Nobody wants the statistical equivalent of emptying your pockets onto the page.
Also, APA style normally omits the leading zero for values that cannot exceed 1. So write p = .022, not p = 0.022. The same applies to correlations and many effect sizes when appropriate.
What should you do with confidence intervals?
If your course asks for confidence intervals, include them. They can be useful because they show the estimated range of the effect rather than forcing the reader to stare at a single value and pretend it is more stable than it is.
For example:
Participants in the intervention group reported lower anxiety scores (M = 12.10, SD = 3.42) than participants in the control group (M = 14.85, SD = 3.88), t(70) = -3.16, p = .002, 95% CI [-4.49, -1.01], d = 0.75.
That is a bit heavier, but still readable. If you have several tests, consider using a table instead of building one enormous sentence that slowly loses the will to live.
Common mistakes when reporting t tests in APA 7
One common mistake is reporting only the p value:
Weak: The result was significant, p = .032.
That does not give enough information. Include the test statistic and degrees of freedom:
Better: The result was significant, t(42) = 2.22, p = .032.
Another common mistake is not reporting the direction of the effect:
Weak: There was a significant difference between the groups, t(58) = 3.41, p = .001.
Which group scored higher? What was the difference actually doing? Do not make the reader infer the whole story from a lonely t value.
A better version is:
Better: Participants in the sleep-deprived condition made more errors (M = 12.35, SD = 3.10) than participants in the rested condition (M = 8.90, SD = 2.84), t(58) = 3.41, p = .001, d = 0.88.
A third mistake is over-explaining the test in the Results section:
Too much: An independent-samples t test was used because the study had two separate groups and the researcher wanted to compare their mean scores.
Usually, your reader does not need that in the Results section unless there is something unusual about the analysis. Save method justification for the Method or analysis plan. The Results section should report what happened, not narrate the statistical admin.
A fourth mistake is using dramatic language:
Too much: The results proved that sleep deprivation destroys memory performance.
A t test rarely proves anything with that level of swagger. Try:
Better: Participants in the sleep-deprived condition recalled fewer words than participants in the rested condition.
Calmer. Cleaner. Less likely to annoy whoever is marking it.
Quick checklist for reporting a t test
Before you submit, check that your write-up includes the following:
The type of comparison is clear.
The groups, conditions, or comparison value are named.
The means and standard deviations are included.
The direction of the difference is stated.
The t value is reported.
The degrees of freedom are in parentheses after t.
The p value is reported correctly.
The effect size is included if required.
Statistical symbols are italicised.
You have not interpreted the result like you are already halfway through the Discussion section.
A simple template you can adapt
For an independent-samples t test:
Participants in [Group 1] had higher/lower [outcome] scores (M = XX.XX, SD = XX.XX) than participants in [Group 2] (M = XX.XX, SD = XX.XX), t(df) = X.XX, p = .XXX, d = X.XX.
For a paired-samples t test:
Participants had higher/lower [outcome] scores in [Condition 1] (M = XX.XX, SD = XX.XX) than in [Condition 2] (M = XX.XX, SD = XX.XX), t(df) = X.XX, p = .XXX, d = X.XX.
For a one-sample t test:
Participants’ [outcome] scores (M = XX.XX, SD = XX.XX) were significantly higher/lower than [comparison value], t(df) = X.XX, p = .XXX, d = X.XX.
If the result is non-significant, replace “were significantly higher/lower” with “did not differ significantly from” or “there was no significant difference between.”
Free article helper
APA t-Test Reporter
Generate a cleaner APA-style t-test sentence for psychology and social science assignments, without manually stitching one together from output tables and quiet despair.
1. Choose your t-test
Pick the type of result you want to report. The fields below will adjust to match.
2. Enter your values
Add the result details from your output. Keep it simple. This is not the place for statistical theatre.
3. Your APA sentence
Copy this into your results section, then check it against your module guidance. Some markers enjoy additional requirements. It gives them something to do.
This helper formats a reporting sentence. It does not check whether you chose the correct test, met the assumptions, or accidentally angered Levene’s test.
Need more than t-tests?
This is the small version. Deliberately.
This free helper covers quick t-test write-ups only. The full Results Reporter in the Formatting Pack helps with common psychology results, including ANOVA, correlation, chi-square, regression, and cleaner APA-style wording across the usual statistical suspects.
Final thought
A good APA t test write-up is not glamorous. It is meant to be clear, complete, and slightly invisible. The reader should understand the result without stopping to untangle your sentence, check your output, or wonder why the Results section has developed a personality disorder.
Report the means. Report the test. Report the p value. Add the effect size. Say which direction the difference went. Then leave. There is dignity in knowing when a statistics sentence has finished.
Got the output but not the wording?
The Original Matter Stats Pack helps turn common psychology statistics into cleaner APA-style results sentences, without making you manually wrestle every t, p, and effect size into place.
Use it when you know what test you ran, but your results section is still looking at you like it wants a fight.
References
American Psychological Association. (2024). Number and statistics guide: APA Style 7th edition.
Lakens, D. (2013). Calculating and reporting effect sizes to facilitate cumulative science: A practical primer for t-tests and ANOVAs. Frontiers in Psychology, 4, Article 863. https://doi.org/10.3389/fpsyg.2013.00863
