Free · no account · nothing leaves your browser

Free statistical tests for medical research

Run the two tests used most in clinical papers, an independent-samples t-test and a chi-square / Fisher exact test, straight from your summary numbers. You get the p-value, the effect size and its 95% confidence interval, and a sentence you can adapt for your results section.

Calculators

Variances

Example: ICU stay in days, malnourished vs well-nourished children.

Which statistical test should I use?

Choose the test from the type of outcome and how the groups are formed. If you are unsure whether a continuous outcome is normally distributed, look at a histogram and run a Shapiro–Wilk test; with skewed data or small samples, use the non-parametric alternative.

Outcome2 independent groupsSame patients twice3 or more groupsRelationship
Continuous, normalIndependent t-test (Welch)Paired t-testOne-way ANOVAPearson correlation, linear regression
Continuous, skewed or ordinalMann–Whitney UWilcoxon signed-rankKruskal–WallisSpearman correlation
Categorical (yes / no)Chi-square, Fisher exactMcNemarChi-square (r × c)Logistic regression, ROC curve

Reporting your results

Journals and the CONSORT and STROBE guidelines ask for more than a p-value. Report the effect size with its 95% confidence interval (mean difference, odds ratio or risk ratio), the test you used and why, and the number analysed in each group. A p-value below 0.05 says the difference is unlikely to be due to chance alone; the confidence interval shows how large, and how clinically important, it may be.

All the tests, on your full dataset

Library Medicine® runs these tests on the raw data you collect, with assumption checks built in:

  • Descriptive statistics
  • Cross-tabs
  • Chi-square
  • Fisher exact
  • t-tests
  • Mann–Whitney
  • Wilcoxon
  • ANOVA
  • Kruskal–Wallis
  • Shapiro–Wilk
  • Correlation
  • Linear regression
  • Logistic regression
  • ROC curves
  • Sample size

From data collection to statistics in one place

Stop copying data between paper forms, spreadsheets and statistics packages. Collect patient data with validated forms, online or offline, then analyse it in the same project and build charts for your report. When you need SPSS, Stata or R, export with variable and value labels in one click.

Discover the medical research data collection software

Frequently asked questions

Which statistical test should I use for my medical study?

It depends on the outcome and the groups: a t-test (or Mann–Whitney U if skewed) compares a continuous outcome between two groups, ANOVA or Kruskal–Wallis compares three or more, a chi-square or Fisher exact test compares proportions, and regression adjusts for confounders.

When should I use Welch’s t-test instead of Student’s t-test?

Welch’s t-test does not assume equal variances in the two groups, so it is the safer default. When the variances really are equal it gives almost the same answer as Student’s t-test.

When should I use Fisher’s exact test instead of chi-square?

Use Fisher’s exact test when any expected cell count in a 2 × 2 table is below 5, which is common in small clinical studies. With larger counts the chi-square test and Fisher’s test agree closely.

Are these calculators really free?

Yes. They run in your browser with no account, and nothing you type is sent to a server. To analyse a full dataset with assumption checks, regression and charts, create a free Library Medicine® account.

Can I export my data to SPSS, Stata or R?

Yes. Library Medicine® exports your data with variable and value labels to SPSS, Stata and R, as well as CSV and Excel.

More free tools for medical researchers

Medical research data collection software, ready in minutes

Design forms, collect patient data online or offline, and analyse it with built-in statistics in one secure workspace for clinicians, students and research teams.