论文部分内容阅读
科研中经常遇到涉及两个因素的实验,若两因素有交互作用,则需采用析因设计、正交设计等设计类型,但这些设计通常需要进行较多次实验。若在理论和专业上可以证明,实验中涉及到的两个因素不存在交互作用,或交互作用对观测结果的影响无统计学意义,并且特定实验条件(通常为各因素不同水平的一种组合)下实验数据的测定误差在专业上允许的范围之内,则可采用无重复实验的随机区组设计、无重复实验的平衡不完全随机区组设计、具有一个重复测量的单因素设计、无重复实验双因素设计和两因素嵌套设计。本文结合实例,重点介绍后两种实验设计类型。
Scientific research often involves experiments involving two factors. If there is interaction between the two factors, design types such as factorial design and orthogonal design are required. However, these designs usually require more experiments. It can be proved theoretically and professionally that there is no interaction between the two factors involved in the experiment or that the interaction has no effect on the observation results and that the specific experimental conditions (usually a combination of different levels of each factor ) The determination of experimental data error within the professional allowable range, you can use non-repeatable randomized block design, non-repeatable experimental incomplete randomized block design, with a single measurement of repeated measurement design, no Repeat experimental two-factor design and two-factor nested design. This article with examples, focusing on the latter two experimental design types.