Efficacy evaluation under type II censored plan with postclassified samples.
The main theme of this paper is efficacy evaluation under Type II censoring. Observations from a clinical trial are often postclassified into two groups in an effort to properly assess the efficacy of an intervention on each group. Thus, the numbers of observed lifetimes in the two groups are random outcomes. Parameter estimation and hypotheses testing under this postclassified setting are discussed. A numerical study is conducted to assess whether the results derived based on fixed sample sizes could be used in the postclassified case. The results would provide insight to practitioners in analyzing data observed under the postclassified case.