当要研究试验的因子较多时 ,完全析因试验的测试次数会多到甚至无法实施的地步 ,或者造成了浪费。本文概述了测试次数仅为完全析因试验的某个分数值 ,即采用部分实施的办法 ,以集中考察主效应及其比较重要的交互效应。着重阐述了两水平试验设计 ,并借此简明地解释清楚部分实施设计的基本方法 ,且以部分实施与完全析因试验作比较分析。对不同设计方法、混淆混杂的构造及其释放分离等均作了探讨。对于三水平试验设计等也作了相应的概述和探讨。 When there are more factors to study, the full factorial test can be more or less practicable or wasteful. This article outlines that the number of tests is only a fraction of the full factorial test, ie a partially implemented approach to focus on the main effects and their more important interaction effects. Emphasis is given to the design of two-level test. In this way, the basic method of partial design implementation is briefly explained and compared with the partial factorial analysis. The different design methods, confused hybrid structure and its release are discussed. For the three levels of experimental design also made a corresponding overview and discussion.