Statistics decision support for clinicians
Pick the right statistical test for your study in two minutes.
Answer a few questions about your outcome and design. Get the test, its assumptions, what to use instead, and the mistakes reviewers catch. In plain language, with clinical examples.
Question 1 of up to 6
What do you want to find out?
- Whether an outcome differs between groups e.g. HbA1c in treated vs control
- How strongly two numeric measurements move together e.g. BMI and blood pressure
- An association adjusted for confounders, or a prediction model e.g. smoking and MI, adjusted for age and sex
- How long until an event happens death, relapse, discharge; some patients censored
- and 3 more
How it works
1
Describe your study
Three to five questions: outcome type, number of groups, paired or not, and whether you need to adjust for confounders.
2
Get the test, with reasons
Why it fits, which assumptions to check, and what to use instead when they fail.
3
Write it up
Pro generates a statistical methods paragraph for your paper, an R or SPSS script, and a reporting checklist.
Browse by outcome
Already know what you're measuring? Go straight to the methods.
What not to do
The chooser warns you at the right moment. A few of the classics:
Letting a normality test decide
In large samples a trivial deviation is 'significant'; in small samples real skew goes undetected. Look at the plot and the sample size, and check Shapiro–Wilk's W rather than its p-value. With about 25+ per group, mild skew is fine for a t-test.
Reporting only p-values
Give the effect size with a 95% confidence interval. It shows direction, size and precision.
Calling an odds ratio a risk ratio
When the outcome affects more than about 10% of patients, the OR overstates the relative risk.
Correlation as proof of agreement
Two devices can correlate at r = 0.95 and still differ by a clinically important amount. Use Bland–Altman.
Pricing
Pro
€10 per month
- Everything in Free
- Methods paragraph generator
- R or SPSS script generator
- Reporting checklist per study design