Abstract
A study of otitis media that requires a test for the comparison of two clustered samples of censored data is described. A method is proposed taking into account the within-subject correlation in the formation of the log-rank statistic.
| Original language | English (US) |
|---|---|
| Pages (from-to) | 328-334 |
| Number of pages | 7 |
| Journal | Biometrics |
| Volume | 52 |
| Issue number | 1 |
| DOIs | |
| State | Published - Mar 1996 |
Keywords
- Log-rank test
- Paired survival times
- Survival analysis
- Within-subject correlation