Paired-samples t test
The paired-samples t test compares two measurements taken on the same cases — before and after, or two conditions — by testing whether the mean of the differences is zero.
Menu: Analyze → Compare Means → Paired-Samples T Test…
When should I use Paired-samples t test?
- Pre-test vs post-test on the same participants.
- Matched pairs.
What are the assumptions?
- The differences are roughly normal (or n is reasonably large).
How do I run it in StatLab?
- Choose Variable 1 (e.g. pre) and Variable 2 (e.g. post).
How do I read the output?
| Output | Meaning |
|---|---|
| Paired Samples Statistics | Mean, N, SD for each measurement. |
| Correlation | How strongly the two measurements go together. |
| Mean, SD, 95% CI of the differences | Variable 1 − Variable 2. |
| t, df, Sig. | Tests whether the mean difference is zero. |
| Cohen's dz | Mean difference / SD of differences. |
Worked example
Compare pre and post in the bundled sample dataset (File → Sample dataset: 150 survey respondents with gender, programme group, age, pre/post test scores, five satisfaction items, income, smoking and education): mean difference (pre − post) = −3.07, t(149) = −6.36, p < .001.
How would I report it in APA style?
Post-test scores (M = 54.45, SD = 12.48) were significantly higher than pre-test scores (M = 51.38, SD = 10.49), t(149) = −6.36, p < .001, dz = 0.52.
Numbers above were computed by StatLab’s validated engine from the sample dataset when this page was built.
What are common mistakes?
- Reading the sign the wrong way: the difference is Variable 1 minus Variable 2.