T-Test Calculator
Enter each group's mean, standard deviation, and sample size to calculate the t-statistic and degrees of freedom.
Last updated: September 5, 2026
Frequently asked questions
What does the t-statistic tell me?
It measures how many standard errors apart the two group means are - the larger |t| is, the less likely the difference between groups happened by chance alone, assuming the null hypothesis (no real difference) is true.
How do I know if my result is statistically significant?
Compare your calculated |t| to a critical value from a t-distribution table, using your degrees of freedom and chosen significance level (commonly 0.05) - if |t| exceeds the critical value, the result is significant at that level.
What's the difference between this and a paired t-test?
This calculator assumes two independent groups (like a treatment group vs. a control group of different people). A paired t-test instead compares two measurements on the same subjects (like before/after) - for that, calculate the mean and standard deviation of the differences, then treat it as a one-sample t-test against zero.
Why does this use "pooled" standard deviation?
Pooling combines both groups' variance into a single weighted estimate, which is the standard approach when you can reasonably assume both groups have similar underlying variance - it uses more of your data than relying on either group's variance alone.
What does it mean if t is negative?
A negative t simply means Group 1's mean is lower than Group 2's mean - the sign reflects direction, not strength. Only the magnitude (|t|) matters for judging how far apart the groups are.
How are degrees of freedom calculated for this test?
For the pooled two-sample t-test, degrees of freedom = n1 + n2 − 2. With 30 people in each group, that's 30 + 30 − 2 = 58 degrees of freedom.
What if my two groups have very different sample sizes?
The pooled standard deviation formula already weights each group's variance by its own sample size, so unequal n1 and n2 are handled correctly - just be aware that if the group variances are also very different, an unequal-variance (Welch's) t-test may be more appropriate than this pooled version.