paired test calculator

What is a paired test?

A paired test compares two measurements that come from the same unit. In practice, this usually means the same person, product, or system measured twice: once before an intervention and once after.

The most common paired test for numeric data is the paired t-test. Instead of comparing raw groups directly, it computes one difference per pair and tests whether the average difference is zero.

When to use a paired t-test

• You have two measurements for each subject (pre/post, left/right, method A vs method B on the same sample).
• Your outcome is continuous (score, weight, blood pressure, time, etc.).
• You want to test whether the average change is statistically different from zero.

If your two groups are unrelated (different people in each group), use an independent samples t-test instead.

How to use this paired test calculator

1. Enter the Before values in Sample A.
2. Enter the matching After values in Sample B.
3. Pick a hypothesis type (two-tailed, right-tailed, or left-tailed).
4. Set your significance level, usually 0.05.
5. Click Calculate Paired Test.

The calculator reports sample size, mean difference, standard deviation of differences, t-statistic, degrees of freedom, p-value, Cohen’s d_z, and a confidence interval for the mean difference.

The formula behind the paired t-test

Step 1: Compute paired differences

For each pair, compute:

di = Afteri - Beforei

Step 2: Compute the test statistic

Let d̄ be the average difference, sd the sample standard deviation of the differences, and n the number of pairs. Then:

t = d̄ / (sd / √n)

The degrees of freedom are df = n - 1.

Step 3: Interpret the p-value

The p-value is computed from the Student’s t distribution. If p < α, you reject the null hypothesis and conclude that the average change is statistically significant.

Assumptions checklist

• Pairs are correctly matched.
• Pairs are independent from one another.
• Differences are roughly normally distributed (especially important for small n).
• No severe data entry or measurement errors.

For larger sample sizes, mild normality violations are less problematic due to the central limit theorem.

How to read the output

• Mean difference: average change (After - Before).
• t statistic: signal relative to noise in paired differences.
• p-value: evidence against a zero-mean change.
• Cohen’s d_z: standardized effect size for paired data.
• 95% CI: plausible range for the true mean difference.

Common mistakes to avoid

• Using unmatched rows (pair 1 in A must match pair 1 in B).
• Running an independent t-test on paired data.
• Interpreting “not significant” as “no effect at all.”
• Ignoring effect size and confidence intervals.

Quick example

Suppose you measure reaction time for the same 12 participants before and after a training program. If the calculated mean difference is negative and significant, it suggests reaction times decreased after training (improvement).

Click Load Example above to try this workflow instantly.

Paired t-test vs alternatives

If paired differences are highly non-normal or include major outliers with a small sample, you may consider a Wilcoxon signed-rank test as a non-parametric alternative.