Talk:Logrank test
 | This article is rated Start-class on Wikipedia's content assessment scale. It is of interest to the following WikiProjects: | ||||||||||
| |||||||||||
Log-rank test, beyond two-variable unweighted model[edit]
I think more complicated versions of log-rank test should be added to the article as well. For example tests with more than two groups could be a good start (page 12 of this lecture note). I'll try to add if I get the time. But if someone could do this, it would be great. Blue down quark (talk) 13:54, 9 April 2018 (UTC)
Nomenclature - logrank test[edit]
The test is named the logrank test, not the log-rank test. Even the Stanford course notes referenced above have it right, as Stanford runs a fine school of statistics. Despite the AMA style guide, the AMA does not get to rename statistical tests any more than the ASA gets to rename body parts. From the defining paper: "These W-scores sum identically to zero for both groups together, and we call the permutational test based on sums of them the logrank test." [1]
Smckinney2718 (talk) 00:09, 16 June 2018 (UTC)
References
- ^ Richard Peto and Julian Peto (1972). "Asymptotically Efficient Rank Invariant Test Procedures". J. R. Statist. Soc. A, 135, Part 2, p. 185