Efficiency of Cox's Model in Estimating Regression Parameters with Grouped Survival Data
研究了分组数据下非参数生存分析（Cox模型）相对于参数分析的效率，避免了部分似然的困难，并允许使用标准似然方法，结果与未分组数据类似但更易理解。
The efficiency of nonparametric survival analysis (Cox, 1972, 1975) relative to parametric analysis is examined for grouped data. This setting avoids the difficulties of partial likelihood and allows standard likelihood methods to apply. The results obtained here are similar to published results for ungrouped data, but are perhaps more easily understood. Some insight is also gained for problems involving time-dependent covariables and for problems with many covariables.