Answers are grounded in this paper and cite their sources.
Small-sample properties of censored-data.ranktests, such as Mantel's logrank test, the Breslow test or the general dass of tests proposed by Peto and Peto (1972, Journal oj the Royal Statistical Society, Series A 135, 185-206), are examined.It is found that the latter dass of tests, in particular, can be substantially nonconservative when applied to small sam pies.The most serious inaccuracies occur in unbalanced trials, when higher event rates are inferred in the smaller sampie.
No takes yet. Share an insight, caveat, or question.
Kellerer et al. (1983) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: