论文部分内容阅读
在农业試驗中,有許多資料需要用百分数(成数)来表示,如不同小麦品种发生黑穗病穗的百分率、某几种药剂施用后害虫为害百分率、蚕豆不同摘心方法的落花落荚百分率等等。对这些試驗,如果直接用百分数資料进行变異量分析,往往不能获得正确的結果,甚至会作出錯誤的結論。变異量分析法适用于所研究的变量呈正态分布的資料。由于百分数資料,一般呈偏斜状态分布,如果资料的百分率低于30%或高于70%,必須先将原資料轉換成反正弦角度—sin~(-1)P~(1/2),然后再进行变異量分析。为了清楚起見,可举一个例子予以說明。
In agricultural experiments, there are many data that need to be expressed as a percentage (eg, number), such as the percentage of smut ears in different wheat cultivars, the percent damage of pests after the application of certain pesticides, and the percentage of falling petioles in different methods of pinching broad beans Wait. For these tests, if we directly analyze variance with the percentage data, we can not get the correct result, and even make the wrong conclusion. Variables analysis is applied to the normally distributed variables of the studied variables. Since the percentage data is generally skewed, if the percentage of the data is less than 30% or higher than 70%, the original data must be converted to arc sine (-sin ~ (-1) P ~ (1/2)), Then analyze the variation. For the sake of clarity, an example can be given for explanation.