Paired Samples T-Test calculator

Compare means of two related groups (pre/post designs) with our online paired samples t-test calculator. Get instant p-values and effect sizes.

What it does

Compares two measurements on the same subjects by analysing the within-pair differences, so each subject serves as their own control.

Cautions

• Assumes the within-pair differences are approximately normal (raw scores need not be); robust with n ≥ 30 pairs.
• Without a control group, a significant change is not evidence the intervention caused it — time trends, regression to the mean, and placebo effects can all confound an uncontrolled before–after comparison.

Alternatives

• Wilcoxon signed-rank — differences are non-normal and n is small (< 30 pairs)
• Sign test — differences are asymmetric and extreme — Wilcoxon needs rough symmetry of differences, the sign test does not

How to read the output

Mean difference and SD of differences

The average change per subject and how much that change varies between subjects. The test is a one-sample t-test on this column of differences. A large mean change with a large SD of differences means the average hides subjects moving in opposite directions. Plot the paired changes before describing it as a uniform effect.

t-statistic and N (pairs)

df is pairs − 1. Only complete pairs count: a subject missing either measurement is dropped entirely. Check the excluded-rows warning. Dropping incomplete pairs is fine when the missingness is unrelated to the change, and quietly biasing when it is not — for example when the sickest patients missed the follow-up.

p-value

How surprising a difference at least this large would be IF the groups really were identical on this measure. A small p means the data would be surprising under that assumption. It is not the probability that the result is due to chance, and not the probability that the groups are the same — those read the conditional backwards. It also answers 'would this be surprising?', never 'is this big enough to matter?': with a large sample a trivial difference returns a small p, and the result looks identical to an important one.

How this calculator is validated · Which statistical test should I use?