Glossary

Log-Rank Test

A non-parametric hypothesis test used in survival analysis to compare the event-time distributions across two or more groups. It accounts for right-censored observations and uses the full follow-up data. The test is most powerful when the proportional-hazards assumption holds.

Definition

A non-parametric hypothesis test used in survival analysis to compare the event-time distributions across two or more groups. It accounts for right-censored observations and uses the full follow-up data. The test is most powerful when the proportional-hazards assumption holds.

Why It Matters

In clinical trials and reliability studies, participants often drop out or have not yet experienced the event by the study's end. The log-rank test handles this censoring directly, making it the standard method for comparing survival curves between treatment and control groups without imposing distributional assumptions on survival times.

Example

In a randomised trial comparing two cancer therapies, researchers track time-to-relapse for 200 patients over five years. Some patients are still relapse-free at the end of follow-up (censored). The log-rank test determines whether the relapse-free survival distributions differ significantly between the two treatment arms.

Related Terms

Software Notes

  • SPSS: Analyze > Survival > Kaplan-Meier, then select "Compare factor levels" with Log Rank
  • R: survdiff(Surv(time, status) ~ group, data = df) from the survival package
  • Stata: sts test group