Table 1 generator – baseline characteristics table with tests and methods text
Paste your data from Excel – one row per participant – and get a publication-ready Table 1 split by group, with suitable tests, SMD and a methods text. Free, no account; your data are neither uploaded nor stored.
More about this tool
What you can do
- Build Table 1 from raw data, split by a grouping variable with up to six groups, optionally with an overall column.
- Show continuous variables as mean ± SD, mean (SD) or median (Q1; Q3) and categorical variables as n (%). In automatic mode the Shapiro-Wilk test in each group chooses between mean and median; you can override it and relabel each variable.
- Add P values with a column naming the test, and the standardised mean difference (SMD) when there are two groups.
- Show yes/no variables in one row and add a “Missing” row when values are missing.
- Get footnotes and a suggested methods text in the table language of your choice (nine languages).
How it works
- Paste the data: first row column names, then one row per person. Columns may be separated by tab (copied directly from Excel), semicolon or comma; decimal comma and point are both accepted; empty cells and “NA” count as missing. Up to 20,000 rows and 80 columns; CSV and TXT files can be loaded.
- Choose the grouping variable. Numbers with more than five distinct values are treated as continuous, everything else as categorical; columns such as “ID” are skipped.
Tests: two groups – Welch t test or Mann–Whitney U test; more than two groups – one-way analysis of variance or Kruskal–Wallis test; categorical variables – Pearson chi-square test, or Fisher’s exact test for 2 × 2 tables with small expected counts. The SMD for proportions follows Austin.
Export
Copy the table straight into Word or Excel, or download it as CSV, SVG, vector PDF (Latin script only) or PNG at 300 or 600 dpi, 120, 170 or 240 mm wide. “Save project” writes a file to continue later.
Your data: everything is calculated in your browser. The data are neither transmitted nor stored; only your settings are remembered on this device. Please do not enter names or dates of birth.
Frequently asked questions
Should Table 1 of a randomised trial contain P values?
CONSORT advises against significance tests of baseline values, because with proper randomisation any differences arise by chance. Describe the groups and use the SMD if needed; the tile reminds you of this when P values are switched on.
Mean ± SD or median (Q1; Q3)?
Mean and standard deviation suit roughly symmetric variables, median and quartiles suit skewed ones such as CRP or length of stay. The Shapiro-Wilk test helps with the choice, but in small groups it has little power.
What does an SMD below 0.1 mean?
Values below about 0.1 are often regarded as well balanced. Unlike the P value, the SMD does not depend on group size.
pubmed.info sets no tracking cookies and builds no user profiles. Visits are counted anonymously (cookieless, shortened IP addresses). Ads from Google appear only according to your consent choice.