论文部分内容阅读
本刊在1979年16卷6期发表了王颉修同志的“对宋哲和著《农药药效试验的设计与分析》一书的若干意见”一文(以下简称“意见”),提出了对显著性检验中多重比较问题的不同看法。认为《分析》一书“无根据的否定了t测验的作用而错误的介绍邓肯(Duncan)氏和哈特利及柯尔斯(Hartley & Keuls) 氏(称Q值法)测验法。”并提出了“差异显著标准应该是统一的。”的意见。因为多重比较问题是科学实验尤其是农药药效试验分析中经常遇到的问题,其方法很多,应用的条件和效果也各不相同,因此,想就一些常用方法的检验功能和判断原则
The Journal published a number of Opinions on Song Zhe and his book “Design and Analysis of Pesticide Efficacy Test” published by Wang Jiexiu in 1979, Volume 16, Issue 6 (hereinafter referred to as “Opinion”), Different views on multiple comparison problems in testing. Arguing that the Analytic book “unfounded” denied the role of the t-test and erroneously introduced Duncan’s and Hartley & Keuls’s (Q-test) methods. Put forward the opinion that “the significant standard of difference should be unified.” Because multiple comparison problems are often encountered in scientific experiments, especially in the analysis of pesticide efficacy tests, there are many ways and different conditions and effects of application. Therefore, to test some commonly used methods of inspection function and judgment principle