Burman Carl-Fredrik, Sonesson Christian
AstraZeneca R & D, SE-431 83 Mölndal, Sweden.
Biometrics. 2006 Sep;62(3):664-9; discussion 670-83. doi: 10.1111/j.1541-0420.2006.00626.x.
Flexible designs allow large modifications of a design during an experiment. In particular, the sample size can be modified in response to interim data or external information. A standard flexible methodology combines such design modifications with a weighted test, which guarantees the type I error level. However, this inference violates basic inference principles. In an example with independent N(mu, 1) observations, the test rejects the null hypothesis of mu < or = 0 while the average of the observations is negative. We conclude that flexible design in its most general form with the corresponding weighted test is not valid. Several possible modifications of the flexible design methodology are discussed with a focus on alternative hypothesis tests.
灵活设计允许在实验过程中对设计进行大幅修改。特别是，可以根据中期数据或外部信息修改样本量。一种标准的灵活方法将这种设计修改与加权检验相结合，这保证了第一类错误水平。然而，这种推断违反了基本的推断原则。在一个具有独立N(μ, 1)观测值的例子中，检验拒绝了μ≤0的原假设，而观测值的平均值为负。我们得出结论，最一般形式的灵活设计及其相应的加权检验是无效的。本文讨论了灵活设计方法的几种可能修改，重点是备择假设检验。
