This online group comparison tool lets you upload Excel data and compare independent groups using Welch’s t-test, one-way ANOVA, Mann–Whitney U, or Kruskal–Wallis tests. It follows the group comparison workflow used in our Cardiovascular Diabetology study, with summary statistics, Holm-adjusted pairwise comparisons, and downloadable plots. For example datasets and cardiovascular context, see the 4D flow MRI E/A reference values, PWV reference values, and the 4D flow MRI research hub.
From our Cardiovascular Diabetology research
Our study compared lean adults, adults with overweight, and adults with overweight and type 2 diabetes, including quantitative CMR and 4D flow MRI measurements. This tool implements the independent-group comparison component: distribution assessment, descriptive statistics, overall tests, and Holm-adjusted pairwise comparisons.
Read the Cardiovascular Diabetology paper · Explore the study and E/A vorticity findings
The current app reports all available pairwise comparisons, including non-significant results. The paper described post-hoc testing after a significant overall result. Correlation, regression, and sample-size analyses from the paper are separate from this tool.
Sample results
Statistical methods
| Analysis | Method | Reporting |
|---|---|---|
| Distribution assessment | Shapiro–Wilk within each group | The parametric branch is used when every available group has a valid test with P > 0.05; otherwise the non-parametric branch is used. |
| Two independent groups | Unpaired Welch’s t-test or Mann–Whitney U (Wilcoxon rank-sum) | Two-sided P-value |
| Three or more groups, parametric | One-way ANOVA; pooled-SD pairwise t-tests | Overall P and Holm-adjusted pairwise P-values |
| Three or more groups, non-parametric | Kruskal–Wallis; Dunn’s pairwise tests | Overall P and Holm-adjusted pairwise P-values; ties are accounted for. |
| Summary statistics | Mean ± SD or median (IQR) | Available sample size for each group and outcome |
| Missing observations | Available-case analysis per outcome | Blank values are excluded only for that outcome; an entirely missing group is omitted from its comparison. |
Holm correction applies to pairwise comparisons within each outcome, not across all uploaded outcomes. ANOVA and pooled-SD comparisons assume equal variances; the app does not run a variance test. A normality test alone does not establish that a method is suitable for a study design.
How to use
- Prepare an Excel workbook (.xlsx). Put column names in the first row, a group identifier in one column, and numeric outcomes in the remaining columns. Each row represents one independent observation. Leave missing values blank.
- Upload and select. Choose the worksheet and group column, then click Start group comparison. Run it again after changing the uploaded data, worksheet, or group column.
- Review the results. Inspect sample sizes, summary statistics, overall and pairwise P-values. Choose an outcome and dot, box, or violin style. The black marker shows the mean or median, according to the analysis branch.
- Adjust and export. The Y-axis range slider adjusts plot padding and comparison spacing. Download the results as Excel, the plot as a vector PDF, or a PNG, JPG, or TIFF at your selected DPI. Images default to 300 DPI. All exports use a 7 × 5 inch canvas; vector PDF does not have a fixed DPI.
Try the example data
Download the workbook or copy the table into Excel, save it as .xlsx, and upload it to the tool. This illustrative dataset has three groups and five parameters. Parameter 1 has no values in group3, so it demonstrates a two-group comparison; Parameters 2–5 demonstrate three-group comparisons.
Standard .xlsx · fixed numeric values · no formulas or macros · no external links · no installation required
| Group | Parameter 1 | Parameter 2 | Parameter 3 | Parameter 4 | Parameter 5 |
|---|---|---|---|---|---|
| group1 | 0.4648755838716816 | 0.6437295418901966 | |||
| group1 | 0.908709941178401 | 0.7483535588858834 | |||
| group1 | 0.9705915858398309 | 0.018272890934245112 | 0.9104562157984408 | 0.0826135808170515 | 0.1455267078778325 |
| group1 | 0.04964066089478103 | 0.30286490029126345 | 0.4688159031157122 | 0.6664236185273948 | 0.02237342435139178 |
| group1 | 0.8180685324002955 | 0.6750140238238842 | 0.8860042236755834 | 0.44530394615564506 | 0.6322919911269349 |
| group1 | 0.8158434847325325 | 0.3663291408893442 | 0.9460063701219228 | 0.5678753383235327 | 0.2506437386349989 |
| group1 | 0.6333846735894711 | 0.891103638667414 | 0.8984458949614433 | 0.9520859457501981 | 0.6194100866577776 |
| group1 | 0.9151166183558452 | 0.004847130224214458 | 0.48857676145214346 | 0.6083847447747776 | 0.9531307314921136 |
| group1 | 0.10733163878823304 | 0.5394895800179798 | 0.9621005832088532 | 0.09179569962191914 | 0.9231984340189282 |
| group1 | 0.8180633979764923 | 0.8888169901066847 | 0.8323921341585531 | 0.26473877532412016 | 0.0845121216444833 |
| group1 | 0.4921932254193093 | 0.15098575859272423 | 0.5589735422737114 | 0.8789711669950628 | 0.8364199121788958 |
| group1 | 0.45512921696511377 | 0.5858860096328141 | 0.18712568396251839 | 0.4284742205687957 | 0.21435270656973826 |
| group1 | 0.8415512507712617 | 0.9001669170542401 | 0.7083796982934135 | 0.2936552652193728 | 0.1887825426755495 |
| group1 | 0.9489579776615563 | 0.06192562521235079 | 0.6892820192348214 | 0.1256580917635347 | 0.7236408359112302 |
| group1 | 0.790632603041082 | 0.41988113027757734 | 0.23604181409058866 | 0.44586676256879343 | 0.6019536529094864 |
| group2 | 0.9124627051054658 | 0.8148248132960589 | 0.07298685010736927 | 0.40902883829613756 | 0.7509015780149392 |
| group2 | 0.33586436898495586 | 0.5148584685480159 | 0.08937590553933206 | 0.3570402490148257 | 0.5869738049126489 |
| group2 | 0.35203628488790006 | 0.23359193198294015 | |||
| group2 | 0.6341044455061633 | 0.900530667345089 | 0.9417365473503332 | 0.8291386160236834 | 0.7866553743075816 |
| group2 | 0.6183534938708343 | 0.6167892649298224 | 0.4809076349458058 | 0.4102024093206864 | 0.740587422731075 |
| group2 | 0.8813440455484045 | 0.003585479970023231 | 0.6011352183655708 | 0.3738500410689266 | 0.9825575041565302 |
| group2 | 0.6826030656504459 | 0.3582385977681679 | 0.9789624423966443 | 0.4662242270847907 | 0.2255988594492334 |
| group2 | 0.2736611601286342 | 0.8871009748852875 | 0.6269580016347198 | 0.1439130775450378 | 0.11245547103616205 |
| group3 | 0.6479541662982037 | 0.04658273031038074 | 0.2835636496840106 | 0.21466983709109644 | |
| group3 | 0.032451665599260715 | 0.7321731043227857 | 0.2574400135828051 | 0.8468842369261802 | |
| group3 | 0.3664521050900579 | 0.2870653920611076 | 0.037002046161028423 | 0.7857742326616056 | |
| group3 | 0.6716528269632233 | 0.01943795880844812 | 0.3932292835280814 | 0.7238984160080825 | |
| group3 | 0.7893708778441599 | 0.1801049326272156 | 0.19945955905578383 | 0.3514162295133004 | |
| group3 | 0.2296407075913477 | 0.9004092941346576 | 0.29445945310777955 | 0.6888109273835885 |
Values are displayed at full precision so the copied table matches the downloadable workbook. The file preserves the uploaded example’s values and missing cells, with formulas replaced by their stored numeric results.
Primary reference
Park et al. Comprehensive CMR evaluation including 4D flow-derived E/A vorticity ratio in overweight adults with and without type 2 diabetes mellitus. Cardiovascular Diabetology 25, 7 (2026). DOI: 10.1186/s12933-025-02999-9.