Your data
Paste a table straight from Excel or Google Sheets, or drop a CSV file. Nothing is uploaded anywhere — the whole calculation runs inside this browser tab.
CSV, TSV or plain text. Files saved in windows-1251 are detected and re-read automatically, so Cyrillic headers do not turn into garbage.
Parsing options
Columns
Result
Save and share
The drawing library is downloaded only when you press the button, and only on this page.
This page has no server side at all. The table you paste is parsed by JavaScript inside your own browser, the chart is drawn on a canvas element on your machine, and the export file is assembled locally. Nothing is uploaded, stored or written to any log. You can disconnect from the network after the page has loaded and everything will still work — which is the simplest way to verify the claim yourself.
Facts and limits of this method
| Test used | Welch's t-test — does not assume equal variances |
|---|---|
| Why Welch | The classic Student version overstates significance when group sizes or spreads differ |
| Degrees of freedom | Welch–Satterthwaite approximation, usually fractional |
| p-value | Exact, from the t-distribution via the regularised incomplete beta function |
| Effect size | Cohen's d, computed with the pooled standard deviation |
| Minimum rows | 2 per group; meaningfully around 15 each |
When it misleads you
- A p-value is not the probability that your hypothesis is true, and it is not the probability the result was a fluke. It is the probability of seeing a difference at least this large if the two groups really had the same mean — a subtly but importantly different statement.
- Statistical significance is not practical importance. With enough rows, a difference of two seconds in a five-minute task becomes highly significant and remains irrelevant. That is why Cohen’s d is reported next to p: d below 0.2 is a negligible effect no matter how small the p-value.
- The test compares means, so it is blind to everything else. Two groups with identical averages and completely different spreads produce a p-value near 1 while being obviously different processes.
- Running the test after peeking at the data, or repeating it on many pairs until something turns significant, invalidates the p-value entirely. The 5% threshold assumes exactly one pre-planned comparison.
How it is calculated
The test statistic is the difference between the two means divided by the standard error of that difference. The larger the difference and the smaller the internal variation, the larger t becomes, and the less plausible it is that the groups share a mean.
Welch’s version is used rather than the classic Student test. Student’s test assumes the two groups have equal variances, and when they do not — which is the normal case with real data of unequal sizes — it produces p-values that are too small and declares differences significant that are not.
The degrees of freedom come from the Welch–Satterthwaite equation and are usually not a whole number, which is a feature rather than an error: the correction adjusts the shape of the reference distribution to the imbalance between the groups.
Cohen’s d divides the difference in means by the pooled standard deviation, giving a unitless measure of how far apart the groups are relative to their internal spread. Conventionally 0.2 is small, 0.5 medium and 0.8 large, and it does not change when you collect more data — unlike the p-value, which shrinks indefinitely.
Questions and answers
What does p < 0.05 actually mean?
That a difference at least this large would occur less than 5% of the time if the two groups genuinely had the same mean. It is a statement about the data under an assumption, not about the probability that your conclusion is right.
Why Welch instead of the standard t-test?
Because Welch does not assume equal variances and loses almost nothing when they happen to be equal. Modern statistical practice recommends it as the default, and it is what this page runs.
My p is significant but the difference looks tiny.
Look at Cohen’s d. A d of 0.1 with a significant p means you have a lot of data and a difference too small to act on. Significance and importance are separate questions.
Are my two columns paired?
This is the unpaired test, which treats the columns as independent groups. If each row is the same subject measured twice — before and after — a paired test is the correct one and this page will understate the effect.
Is my data uploaded?
No. The statistic, the degrees of freedom and the p-value are all computed locally.