Fractional factorial design
Fractional factorial design
复制标题
DOI：
10.1002/wics.27
复制
发表时间：
2009-09-01
影响因子：
3.2
通讯作者：
Mason, Robert L.
中科院分区：
文献类型：
--
作者：
Gunst, Richard F.;Mason, Robert L.
部分析因设计是有效探索几个可控因素对兴趣反应的影响的最重要的统计学贡献之一。分数阶乘在农业、工业和医学研究等领域的实验中得到了广泛的应用。分数阶乘的一个关键特征是统计特性在试验之前就已知，这是更多用于减小试验规模的特别方法所不具备的。因此，实验者可以研究能够以最少的成本、最短的时间或最有效的资源利用来实现实验目标的替代方案。有时，一旦知道设计的统计特性，实验者可能决定不按原计划进行实验。本文重点介绍了部分析因设计的基本概念、设计策略和统计性质。(C)2009年John Wiley父子公司
Fractional factorial designs are among the most important statistical contributions to the efficient exploration of the effects of several controllable factors on a response of interest. Fractional factorials are widely used in experiments in fields as diverse as agriculture, industry, and medical research. A key feature of fractional factorials that is not shared by more ad hoc methods for reducing the size of experiments is that the statistical properties are known in advance of experimentation. Consequently, an experimenter can investigate alternatives that enable the goals of the experiment to be met with the least cost, shortest time, or most effective use of resources. On occasion, an experimenter might decide not to conduct an experiment as originally planned once the statistical properties of the design are known. This article highlights the fundamental concepts, design strategies, and statistical properties of fractional factorial designs.. (C) 2009 John Wiley & Sons, Inc.