What this is for: Two-group mean comparisons — Student (equal variance), Welch (unequal variance) and paired sheets, with effect size and charts.
The formula: t = (mean1 − mean2) ÷ standard error of the difference
How to use:
- Choose the section that matches your design: Student for independent groups with similar variances, Welch for independent groups with unequal variances, or Paired for matched before/after data.
- Paste the values in that section, one value per line in each group box.
- For the paired test, keep both boxes in the same pair order — row 1 in each box must be the same subject or sample.
- Read n, mean, SD and SEM per group, then the mean difference, t, df, two-tailed p-value, Cohen's d (dz for paired) and significance stars.
- In Chart Settings, choose whether the bracket over each means chart shows stars or the exact p-value.
Unpaired (Student) t-Test — two independent groups, equal variance
Two INDEPENDENT groups, equal-variance (Student) t-test. Paste one value per line in each group. Stars: * p<0.05, ** p<0.01, *** p<0.001.
Welch t-Test — two independent groups, unequal variances
Two INDEPENDENT groups with UNEQUAL variances (Welch's t-test) — the safer default when group SDs differ. df uses the Welch–Satterthwaite equation.
Paired t-Test — each row is one subject measured twice
PAIRED data: row i of Group 1 pairs with row i of Group 2 (e.g. before/after). The test runs on the per-row differences.
Chart Settings
Each chart shows group means ± SEM with a significance bracket over the two bars, labelled from that sheet's own two-tailed p-value (single comparison, so the adjusted p equals the raw p).
Download the t-Tests (Student, Welch, Paired) workbook (.xlsx)