Applied Multivariate Statistics for the Social Sciences
Applied Multivariate Statistics for the Social Sciences
复制标题
DOI：
10.1111/1467-9884.00369_23
复制
发表时间：
2003-10
期刊：
影响因子：
--
通讯作者：
Lada Mitchell
中科院分区：
文献类型：
--
作者：
Lada Mitchell
森讨论了如何处理结转。一种方法是有一个“合适的”清洗期。另一种方法是在存在残留的情况下估计治疗效果，反之亦然。第十章给出了结转问题的数学处理方法。第三章讨论了正态数据下的经典AB-BA设计。在这篇文章中，他讨论了各种分析方法：首先忽略周期效应;然后使用双样本t方法调整周期效应;还有希尔斯和阿米蒂奇（1979）倡导的方法。本章以软件程序（在本例中使用Genstat和S-PLUS编写）结束（正如许多人所做的那样）。第四章讨论了数据变换、模型估计的非参数方法和交叉试验中的二进制数据问题。Kenward和Jones（1987）提出了二进制数据的对数线性模型; Senn使用Genstat统计计算机包对此进行了说明。他们的模型考虑了治疗和周期效应。
Senn discusses what to do with carry-over. One method is to have a'suitable'wash-out period. Another is to estimate the treatment effect in the presence of carry-over and vice versa. Mathematical approaches to the problem of carry-over are given in Chapter 10.The classic AB-BA design with normal data is discussed in Chapter 3. In it he discusses various methods of analysis: first by ignoring the effect of period; then adjusting for the period effect by using a two-sample t-approach; also the method advocated by Hills and Armitage (1979). The chapter ends (as many do) with software programs (written in both Genstat and S-PLUS in this instance). Chapter 4 discusses data transformations, nonpara-metric approaches to model estimation and the problem of binary data in crossover trials. A log-linear model for binary data was proposed by Kenward and Jones (1987); Senn illustrates this by using the Genstat statistical computer package. Their model allows for both treatment and period effects.