The t test is one of the most common analyses in student theses, and one of the most often reported incorrectly. Supervisors and examiners look for the same handful of elements every time: the test you used, the descriptive statistics for each group or time point, the t statistic with its degrees of freedom, the exact p-value, an effect size and, ideally, a confidence interval. This guide shows you exactly how to report an independent samples t test and a paired samples t test in the style of the Publication Manual of the American Psychological Association, 7th edition (APA, 2020), with ready-to-adapt example sentences and a table template.
If you have not run the test yet, you can use the free independent samples t test calculator or the paired samples t test calculator. Both produce SPSS-style output and an APA paragraph that you can check against the rules below.
What an APA t test result must include
APA style does not prescribe a single sentence, but a complete report of a t test normally contains the following elements:
- The name of the test (for example, independent samples t test, Welch’s t test or paired samples t test) and what was compared.
- Descriptive statistics: the mean (M) and standard deviation (SD) for each group or condition, and the sample sizes if they are not reported elsewhere.
- The test statistic: t(df) = value, reported to two decimal places.
- The exact p-value to two or three decimal places (for example, p = .031). Values below .001 are written as p < .001.
- An effect size, usually Cohen’s d (or Hedges’ g for small samples).
- A confidence interval, usually the 95% CI of the mean difference, in square brackets.
- A plain-language statement of the direction of the result (which group scored higher).
APA 7 recommends reporting exact p-values rather than only stating p < .05, and it encourages reporting effect sizes and confidence intervals for all primary results (APA, 2020). Statistical symbols that are Latin letters, such as t, p, M, SD, d and N, are italicised. Do not use a leading zero for values that cannot exceed 1, such as p-values and correlations (write .031, not 0.031), but do use one for values that can, such as Cohen’s d (0.65).
Reporting an independent samples t test
An independent samples t test compares the means of two separate groups, for example male and female students or an experimental and a control group. Before writing the result, check two things in your output.
Step 1: Decide between Student’s and Welch’s t test
Student’s t test assumes that the two groups have equal variances; Welch’s t test does not. SPSS shows Levene’s test and two rows of results: ‘Equal variances assumed’ and ‘Equal variances not assumed’. The traditional rule is to use the first row when Levene’s test is not significant (p ≥ .05) and the second row when it is. Many methodologists now recommend reporting Welch’s t test by default, because it performs as well as Student’s test when variances are equal and much better when they are not (Delacre et al., 2017). Whichever you choose, say so in the text. Welch’s test has fractional degrees of freedom, which you report to two decimal places, for example t(37.42) = 2.15.
Step 2: Collect the numbers
From your output, note each group’s n, M and SD, the t value, the degrees of freedom, the two-tailed p-value, the mean difference with its 95% CI, and Cohen’s d. Our calculator gives all of these in one place.
Step 3: Write the sentence
A clear structure is: purpose, then result, then statistics in parentheses or after a comma. For example:
An independent samples t test was conducted to compare exam scores between students who used the new revision app and those who did not. Students who used the app (M = 74.20, SD = 8.10, n = 32) scored significantly higher than those who did not (M = 68.90, SD = 8.60, n = 30), t(60) = 2.50, p = .015, d = 0.64, 95% CI of the difference [1.06, 9.54].
If the result is not significant, report it in exactly the same way and avoid saying the groups are ‘the same’ or ‘equal’. A non-significant result means the data did not provide enough evidence of a difference:
There was no statistically significant difference in anxiety scores between first-year (M = 21.40, SD = 5.30) and final-year students (M = 20.10, SD = 6.00), t(98) = 1.15, p = .253, d = 0.23, 95% CI [−0.95, 3.55].
When you used Welch’s test, add a short note: ‘Because Levene’s test indicated unequal variances, F = 6.21, p = .015, Welch’s t test is reported.’
Reporting a paired samples t test
A paired samples t test compares two measurements from the same participants, such as pre-test and post-test scores. The structure is the same, but the descriptive statistics refer to the two time points and the degrees of freedom are n − 1, where n is the number of pairs.
A paired samples t test showed that reading comprehension scores increased significantly from the pre-test (M = 64.70, SD = 6.10) to the post-test (M = 68.60, SD = 5.46), t(9) = 4.99, p < .001, dz = 1.58, 95% CI of the mean difference [2.13, 5.67].
For paired designs, the most common effect size is dz, the mean difference divided by the standard deviation of the differences. Some authors prefer an effect size based on the average standard deviation of the two measurements, which is closer to the between-groups d; Lakens (2013) explains the differences. State which one you used.
Presenting t test results in a table
When you compare several outcome variables, a table is clearer than many sentences. APA 7 tables have a number in bold, an italic title, horizontal rules only, and notes below. A typical layout is:
| Variable | Group 1 M (SD) | Group 2 M (SD) | t(df) | p | d |
|---|---|---|---|---|---|
| Motivation | 3.84 (0.61) | 3.52 (0.70) | 2.67 (118) | .009 | 0.49 |
| Self-efficacy | 3.95 (0.55) | 3.88 (0.58) | 0.68 (118) | .499 | 0.12 |
| Satisfaction | 4.10 (0.49) | 3.71 (0.66) | 3.67 (118) | < .001 | 0.67 |
Below the table, add a note that defines abbreviations and the type of test, for example: ‘Note. Group 1 = online learners (n = 60); Group 2 = classroom learners (n = 60). Independent samples t tests, two-tailed.’ If you report the same statistics in a table, you do not need to repeat every number in the text; summarise the pattern instead.
Checking assumptions and reporting them
Examiners often ask whether the assumptions of the t test were checked. In the results or method section, briefly report:
- Normality: for each group (independent t test) or for the difference scores (paired t test). Inspect histograms or Q–Q plots and skewness and kurtosis; the Shapiro–Wilk test can be reported but is very sensitive in large samples (Ghasemi & Zahediasl, 2012). The t test is fairly robust to moderate non-normality, especially with 30 or more participants per group.
- Outliers: state how you identified and handled them.
- Homogeneity of variance: Levene’s test for the independent t test, as above.
If your data clearly violate normality in a small sample, report the non-parametric alternative instead: the Mann-Whitney U test for independent groups or the Wilcoxon signed-rank test for paired data. You can check skewness and kurtosis quickly with the descriptive statistics calculator.
Interpreting the effect size
Cohen (1988) suggested that d values of about 0.20, 0.50 and 0.80 can be described as small, medium and large when there is no better basis for comparison. These are rough conventions, not strict cut-offs; where possible, compare your effect with previous studies in your field. Write the interpretation in words, for example ‘a medium-sized effect’. Our Cohen’s d calculator also gives Hedges’ g and a confidence interval for d. See also our guide to interpreting effect sizes.
Common mistakes to avoid
- Writing p = .000. SPSS displays .000 when p is very small; report it as p < .001.
- Reporting only ‘significant’ or ‘p < .05’ without the t value, degrees of freedom and exact p.
- Forgetting the direction: always say which group or time point had the higher mean.
- Mixing up the degrees of freedom: n₁ + n₂ − 2 for Student’s independent t test; the fractional Welch value when equal variances are not assumed; n − 1 for paired data.
- Using a one-tailed test without having stated a directional hypothesis before collecting the data.
- Running many t tests instead of an ANOVA when you have three or more groups, which inflates the Type I error rate. See how to report ANOVA in APA.
- Leaving out the effect size, which APA 7 expects for primary results.
Quick checklist
- Test named, with Student’s or Welch’s version stated.
- M and SD for each group or time point.
- t(df) to two decimals.
- Exact p-value (or p < .001), without a leading zero.
- Effect size with interpretation.
- 95% CI of the mean difference.
- Assumptions briefly reported.
Roman Urdu mein khulasa
t test ko APA 7 mein report karte waqt ye cheezein zaroor likhein: test ka naam, dono groups ka mean (M) aur SD, t(df) ki value, exact p-value (jaise p = .015; .000 ke bajaye p < .001), effect size Cohen's d aur 95% confidence interval. Agar Levene's test significant ho to Welch t test report karein. Result ka direction bhi likhein, yani kis group ka score zyada tha. Non-significant result ko bhi isi tarah poori statistics ke saath likhein.
Frequently asked questions
How do I write p = .000 in APA?
Never write p = .000. When SPSS shows .000, the p-value is smaller than .0005, so report it as p < .001.
Should I report Student’s or Welch’s t test?
Report Welch’s t test when Levene’s test is significant. Many methodologists recommend Welch’s test by default because it performs well whether or not variances are equal (Delacre et al., 2017).
Do I need to report an effect size for a t test?
Yes. APA 7 recommends effect sizes for primary results. Cohen’s d is the usual choice; for paired data, report d_z or state which version you used.
How many decimal places should I use?
Report t, M, SD and d to two decimal places and p-values to two or three decimal places (APA, 2020).
Can I report t test results only in a table?
Yes. If a table gives the full statistics, summarise the main pattern in the text instead of repeating every number.
Related calculators
- Independent Samples t Test Calculator
- Paired Samples t Test Calculator
- Cohen’s d Effect Size Calculator
- Test chooser: which statistical test should I use?
References
- American Psychological Association. (2020). Publication manual of the American Psychological Association (7th ed.). https://doi.org/10.1037/0000165-000
- Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Lawrence Erlbaum Associates.
- Delacre, M., Lakens, D., & Leys, C. (2017). Why psychologists should by default use Welch’s t-test instead of Student’s t-test. International Review of Social Psychology, 30(1), 92–101. https://doi.org/10.5334/irsp.82
- Field, A. (2018). Discovering statistics using IBM SPSS Statistics (5th ed.). SAGE.
- Ghasemi, A., & Zahediasl, S. (2012). Normality tests for statistical analysis: A guide for non-statisticians. International Journal of Endocrinology and Metabolism, 10(2), 486–489. https://doi.org/10.5812/ijem.3505
- Lakens, D. (2013). Calculating and reporting effect sizes to facilitate cumulative science: A practical primer for t-tests and ANOVAs. Frontiers in Psychology, 4, 863. https://doi.org/10.3389/fpsyg.2013.00863
Written and maintained by the ScholarCalc team. Last reviewed: November 2026.