论文部分内容阅读
在由样本推断总体的假设检验中,有参数统计和非参数统计两大类方法。如应用不当。会降低检验的灵敏度,甚至发生错误。以前认为非参数统计有两大优点:一是总体可为任意分布;二是计算方法简便。现在有了计算器和计算机,这第二个优点就失去了价值,反而暴露出其弊病:很多统计学家对各种非参数统计方法导出了许多简易计算公式和检验界值表,因此统计教材和参考书既要介绍参数统计的公式和检验界值表,又要介绍非参数
In the hypothesis testing of population inference from the sample, there are two kinds of methods: parametric statistics and non-parametric statistics. If not applied properly. Will reduce the sensitivity of the test, and even errors. In the past, non-parametric statistics were considered to have two major advantages: First, it can be arbitrarily distributed; second, the calculation method is simple. Now that there are calculators and computers, this second advantage loses its value, but it exposes its ills: Many statisticians have derived many simple calculation formulas and test threshold tables for various non-parametric statistical methods. The reference book should introduce both the parameter statistics formula and the test threshold table, as well as the non-parametric ones.