目录

1、参数 (parameter)

参数是用于描述总体特征的数值,我们可以从一次测量中获得参数,也可以从一系列对总体的测量中推论得到。

Parameters are numerical values used to describe the overall characteristics, which can be obtained from a single measurement or inferred from a series of measurements of the population.

2、统计量(statistic)

统计量是用于描述样本特征的数值,我们可以从一次测量中获得统计量,也可以从一系列对样本的测量中推论得到。

A statistic is a numerical value used to describe the characteristics of a sample, which can be obtained from a single measurement or inferred from a series of measurements taken on the sample.


可以看到对于参数和统计量的定义有着相似的形式,但是这二者之间同样有着差异。 在对一组样本的测量过程中,统计量是确定的,但是当我们更换样本再次进行测量,统计量的值就可能改变,也就是说,统计量的值是不定的,会随着样本的选取而变化,而参数则是一个固定的值。 在研究中,我们常常用样本统计量来估计总体参数,但样本作为总体的子集,并不能完全等同于总体,在样本统计量和总体参数之间多少存在差异。

There are similar forms of definitions for parameters and statistics, but there are also differences between the two.

In the measurement process of a group of samples, the statistic is fixed, but when we replace the sample and measure again, the value of the statistic may change. That is to say, the value of the statistic is indefinite and will change with the selection of the sample, while the parameter is a fixed value.

In research, we often use sample statistics to estimate population parameters, but as a subset of the population, samples cannot be completely equivalent to the population, and there may be some differences between sample statistics and population parameters.


3、取样误差(sampling error)

取样误差是指样本统计量和样本所对应的总体参数之间的差异。

Sampling error refers to the difference between sample statistics and the population parameters corresponding to the sample.


我们并不能彻底避免取样误差的出现,因为样本和总体之间总是存在差异,基于有限的信息做出的推论也许并不准确,而我们能做的只有运用正确的方法尽量减少取样误差,或是将其保持在研究允许的范围内。

We cannot completely avoid the occurrence of sampling errors, as there are always differences between the sample and the population. Inferences based on limited information may not be accurate, and all we can do is use the correct methods to minimize sampling errors or keep them within the allowable range of the study.