The log-rank test compares the survival curves of two or more groups.
Why it matters
It tells whether differences between curves are larger than chance would give.
Key ideas
- H0: the groups have the same survival over time
- Compares observed and expected events at each event time
- Result: χ², df = groups − 1, p
- Works best if the curves do not cross
- Does not give the size of the difference (use Cox for HR)
- Needs independent groups
Real-life research example
Does survival differ between treatment A and B? A log-rank test gives p = 0.02.
When to use
- Compare survival curves between groups
When NOT to use
- Don't use when curves clearly cross
- Don't report only p without showing the curves
Mini example
Group A: 4 deaths observed, 6.2 expected; Group B: 8 observed, 5.8 expected
Quick takeaway
Log-rank asks: are the curves the same?
Try it yourself
Use the Kaplan–Meier tool with a group variable. Compare the two curves and read the log-rank p-value.
Model answer
A small p suggests the curves differ. Always look at the plot to see which group does better and whether the curves cross.