零配比正交试验设计与结果分析  

Design and Analyses of Orthogonale Test with Null Ratio Factor

作　　者：杨先泉[1] 任正隆[1] 

机构地区：[1]四川农业大学植物遗传与育种实验室,四川雅安625014

出　　处：《生物数学学报》2006年第2期291-296,共6页Journal of Biomathematics

基　　金：四川省教育厅科技基金重点资助项目

摘　　要：发现了正交配比试验的一种特殊情况-零配比正交试验,极差与平方和特征分析表明:常规正交(配比)试验极差与方差分析法均不能直接用于零配比正交试验结果分析．推导了3因素、3水平、无互作零配比正交试验极差估计与平方和分解校正方法,可有效提高试验结果分析的可靠性．The characteristics of a special orthogonale test, null ratio orthogonale test, are investigated in this paper. It is suggested that if there has a null ratio factor, the orthogonality will lose partly. So the conventional analysis methods of range and variance will be inept for this type of experiment. Remediation methods, both for range and variance analyses, were deduced, based on an example orthogonale test with 3 factors, 3 levels and without interactions between factors.

关 键 词：正交试验 零配比 极差分析 方差分析 

分 类 号：O212.6[理学—概率论与数理统计] Q-332[理学—数学]