Differential analysis & tests / Bar chart with group tests
Your data stays in your browser: files are read and processed on your device, never uploaded to any server, and cleared when you close the page. Free, no sign-in.
Bar height is the group mean; error bars show the SD (spread of the data), the SE or the 95% CI (uncertainty of the mean) — state which in your figure legend. Points are the raw data.
With more than two groups an overall test checks whether any group differs; pairwise P values are then adjusted for multiple testing with the method you choose, and the figure shows adjusted P (stars: * < 0.05, ** < 0.01, *** < 0.001, **** < 0.0001, ns = not significant).
A bar chart shows only the mean and hides the shape of the distribution; with small or skewed samples also show the points or a box plot. For compact letters use the compact-letter bar chart.
Use the rank-sum test for skewed data or small samples; t-tests suit roughly normal data (Welch’s version does not assume equal variances). “Each group vs a reference” makes fewer comparisons, so the correction is milder.
Kruskal & Wallis (1952) J Am Stat Assoc 47:583; Mann & Whitney (1947) Ann Math Stat 18:50; Welch (1947) Biometrika 34:28. Statistics match scipy.stats (kruskal, f_oneway, mannwhitneyu, ttest_ind); adjustment matches R p.adjust. The 95% CI is mean ± t(0.975, n−1) × SD/√n.
Up to 100,000 rows; the number of pairwise comparisons grows with the square of the number of groups.
Send us your data and research question and you will receive a written plan within 1 working day: analysis steps, parameter rationale, deliverables and timeline. Quoted per project.