The log-rank test is used to:

The log-rank test is used to:

• A. Estimate the mean time to event in a single group.
• B. Assess how well a model predicts outcomes.
• C. Evaluate the calibration of survival models.
• D. Compare survival curves between groups.

Answer: (D)

The main idea is to determine whether survival experiences differ between groups by comparing entire survival curves over the follow-up period. The log-rank test does this by using the observed number of events at each event time and comparing it to the number expected if the groups had the same survival, summing these differences across all event times while properly handling censored data. Under the null hypothesis of no difference, the test statistic follows a chi-square distribution, so a significant result suggests the survival curves are different. This isn’t about estimating a single summary like the mean time to event in one group, nor is it about predicting individual outcomes or checking calibration of predictions. It specifically assesses whether the overall survival experience differs between groups.

The main idea is to determine whether survival experiences differ between groups by comparing entire survival curves over the follow-up period. The log-rank test does this by using the observed number of events at each event time and comparing it to the number expected if the groups had the same survival, summing these differences across all event times while properly handling censored data. Under the null hypothesis of no difference, the test statistic follows a chi-square distribution, so a significant result suggests the survival curves are different.

This isn’t about estimating a single summary like the mean time to event in one group, nor is it about predicting individual outcomes or checking calibration of predictions. It specifically assesses whether the overall survival experience differs between groups.