Compare before-and-after measurements on the same items. Paste your two columns, get the mean difference, t-statistic, p-value, 95% CI, and Cohen's d — all client-side.
Paste your before and after columns and press Calculate to run the paired t-test.
Watch: 1-Sample, Paired t-tests and Tests for Equal Variances
After a kaizen you measured the same 10 machines before and after. Enter the lab to see whether the improvement is real.
Enter a before value and an after value for each item (machine, patient, workstation). The test works on the differences, so the pairing must be exact — row 1 before corresponds to row 1 after.
The calculator finds the before-to-after difference for each pair, averages those differences, and measures how much the change varies from pair to pair.
It checks whether that average change is big compared to the natural pair-to-pair variation. You get a p-value, a 95% confidence range for the true change, and a practical effect-size score.
The paired t-test is the correct tool for before-and-after comparisons when the same items are measured twice. By working on the differences it removes item-to-item variation and gives a sharper, more powerful test than comparing two independent groups.
The paired t-test tests whether the mean of the differences between paired observations is significantly different from zero. By pairing observations (the same machine before and after a change), you remove between-item variability from the error term and focus entirely on whether the process genuinely changed. It is more powerful than a 2-sample t-test when natural pairing exists.
Because the same items are measured twice (before and after), the test looks only at each item's change rather than comparing two separate groups — which cancels out item-to-item differences and makes small improvements easy to see. It works out the average change, checks how consistent those changes are, and turns that into a t-value and a p-value. A p-value below 0.05 means the change is real. The confidence interval shows how big the change is likely to be, and for metrics like cycle time or defect rate a drop is the improvement you want.
Ten operators have their output measured before and after a training programme. Some improved a lot, some a little, one got slightly worse. Looking at results individually is confusing.
A paired t-test compares each operator to themselves — before vs. after — removing the natural skill differences between people. Average improvement: 4.2 units/hour. p = 0.01. The training worked.
What to do with this: Roll out the training to remaining operators. By comparing each person to themselves, you remove the noise and get a clean answer about whether the training itself made the difference.
Kaizen events, process changes, and training interventions all produce before-and-after data. The paired t-test gives you a rigorous, defensible answer to "did it actually work?" — replacing guesswork with statistical evidence and a concrete estimate of the improvement size and its uncertainty.
Use the paired t-test in the Control phase of DMAIC to confirm an improvement has held, or in the Measure/Analyse phase to verify that a change to one group of items had a consistent effect. The key criterion: every before measurement must correspond to a specific after measurement on the same item.
The paired t-test is the first choice for verifying continuous improvement results on repeated items. It gives you both the significance (p-value) and the practical size of the improvement (mean difference and CI), so you can report both statistical and operational value in your DMAIC tollgate.
If the test is significant, document the improvement in your control plan with the measured before and after values as evidence. Calculate the improvement's financial value (e.g. time saved × cost per hour) for the project report. Monitor the process with a run chart or control chart to ensure the improvement is sustained over time. If the result is not significant, investigate whether the change was implemented consistently across all items, or whether a larger sample is needed.
Combine the paired t-test with run charts, control charts, and a DMAIC project structure to verify and sustain improvements.
A paired t-test is used when the same items (machines, patients, stations) are measured twice — before and after, or under two conditions. Because each pair shares variability, the test eliminates that source of noise and is more powerful than a 2-sample test. A 2-sample (independent) t-test is used when the two groups are completely separate, with no natural pairing between observations.
Data is paired when there is a natural one-to-one correspondence between each before and after value — the same machine, the same patient, the same workstation. Examples: cycle time on 10 machines before and after a process change; blood pressure in 20 patients before and after treatment; inspection times for the same parts under two methods. If the items in the two groups are different, use a 2-sample t-test.
The paired t-test assumes: (1) observations are independent of each other (one pair does not influence another); (2) the differences (after minus before) are approximately normally distributed — for small samples use a normality test; (3) there are no extreme outliers in the differences. Violations of normality with small samples suggest using a nonparametric alternative such as the Wilcoxon signed-rank test.
The minimum recommended is about 10–15 pairs, though the test works with fewer if the differences are normally distributed. The required sample size depends on the expected mean difference, the standard deviation of differences, and the desired power (typically 80%). For a kaizen study where you expect a 10% improvement with moderate variability, 20–30 pairs typically gives adequate power. Use a sample size calculator to plan your study.
A significant paired t-test confirms your improvement worked. Structured training shows you how to sustain it and replicate it elsewhere.
More calculators for statistics, quality, and improvement.
Want to verify and sustain kaizen improvements inside a full DMAIC project? The Black Belt course covers advanced statistics, hypothesis testing, and the full analytical toolkit.
View Black Belt →