t-Test Calculator - One-Sample, Two-Sample (Welch) & Paired t-Tests
t-Test Calculator
Compare means with one-sample, two-sample and paired t-tests
Three Tests
Compare one sample to a known value, two independent groups (Welch or pooled), or paired before-and-after measurements.
Exact p-Values
Tail probabilities are computed to near machine precision, checked against closed forms and published critical values.
Interval and Effect Size
See the confidence interval for the mean or difference and Cohen's d, so you know how large an effect is, not just whether it is significant.
Full Privacy
Your data never leaves your browser.
Choosing and Reading a t-Test
Use a one-sample test to compare a sample mean with a target, a two-sample test for two separate groups, and a paired test when each value in one list matches a value in the other, such as the same people measured twice. Welch's version, the default here, does not assume the groups have equal variance and is the safer choice; tick the box for the classic pooled test.
The p-value is the probability of seeing a difference at least this large if there were really no effect. A p-value below your significance level α (usually 0.05) is called statistically significant. It does not measure the size or importance of an effect, which is why the confidence interval and Cohen's d are shown too. The test assumes roughly normal data, which matters most for small samples.
Key Takeaways
- Sensible default: Welch's test is used unless you ask for pooled variance.
- One or two tails: Choose a two-sided or either one-sided alternative.
- Sample data loaded: Each test opens with a worked example you can overwrite.