6.3.1 Non-probability sampling: commonly used sampling methods 非概率抽样：常用抽取方法在线视频

非概率抽样
Non-probability sampling

我们主要给大家介绍
We will mainly introduce you

以下的三种抽样方式
the following three sampling methods

分别是方便抽样
They are convenience sampling

滚雪球抽样以及定额抽样
snowball sampling and quota sampling

方便抽样就是从便利的角度
Convenience sampling considers sampling acquisition

来考虑获取样本
from a convenience point of view

样本的选取
In sample selection

主要是让调查人员来决定
it is mainly the investigator that decides

这个样本单位
the sampling unit

那么像我们的很多同学
Many of our students may

展开了这个社会调查
conduct social investigations

往往都是采取的这个
They often adopt

方便抽样的方式
the convenience sampling method

比如说有些同学说
For example, some students may want to

我要想调查一下我们学校的学生
investigate the satisfaction of the students

对我们学校这个食堂这个满意度
about the canteen of our school

对我们食堂的这个是否满意
to see how the canteen satisfies the students

这样的一个调查
To conduct such an investigation

那么往往的就到食堂的门口
you may stand at the gate of the canteen

站在那里去发问卷
And distribute questionnaires

那么有些同学会在
Then, some students may

调查报告当中说
say in the investigation report that

我是随即[A1]抽取这个样本单位的
it was by random that I chose this sampling unit

但事实上不是
In fact, it was not the case

你要注意你这个时候
You should pay attention to the fact

是采取的是方便抽样的
you adopted a convenience sampling method

这样一个抽样方式
Such a sampling method

其实就是为了你方便
is just for your convenience

比较在食堂门口人来人往
People are coming and going at the canteen gate

有很多人
There are many people

你比较方便你展开调查
so it’s convenient for you to conduct investigation

还有
Another example

有些同学比如说
Some students may want to

对于某种商品的这些性质的调查
investigate the properties of a commodity

使用情况的调查
and its use

往往可能到一些商场的门口
They may stand at the gates of some shopping malls

进行一些拦截式的访问
and do intercept interviews

还有比如说到人流比较多的
You may conduct such investigation

大街上去展开这个调查
on the street

这些都属于这个方便抽样
They are all called convenience sampling

那么方便抽样
Convenience sampling

我们说是所有的这个抽样方法当中
of all the sampling methods

花费相对是比较少的
costs less than others

并且你的这个抽样单位
and the sampling unit

是可以接近的
is accessible

并且容易测量的
and easy to measure

但是我们说
But we should say

尽管有这样一些优点
though having these advantages

但是我们说它这种抽样方式
this sampling method still

还是有着严重的缺陷
has serious shortcomings

大家想一想你到食堂门口
Just think, you stand at the gate of a canteen

到大街上
on a street

去做拦截式的这种访问
to conduct such intercept interviews

你这肯定不能保证随机性
you can't guarantee randomness

你的研究对象当中
among your research objects

不是所有的人
Not all of them

比如说都会到食堂去吃饭
will go to the canteen to eat

也不是所有的人
Nor will all of them

可能都会在你调查那个时候去逛街
go shopping at the time of your investigation

所以这样的话你就保证不了
So you can't guarantee that

这个样本的这样一个代表性
This is a representative sample

另外大家还有一个问题
And there is still another problem

比如说你到食堂门口去
You go to the canteen gate

去这个拦截式访问的时候
to conduct such intercept interview

你肯定是挑你觉得比较
You must be picking people that you think are

应该是比较顺眼的
pleasing to the eyes

或者你觉得比较好相处的
or easy to get along with

倾向于会回答你这个问题的人
and may answer your questions

进行这个拦截式的访问
In your intercept interview

比如说有些人看起来很严肃
if someone looks serious

或者说很凶
or rather fierce

那你就不会找他来展开这个调查
then you will not ask him questions for investigation

所以这些都是这个方便抽样当中
In convenience sampling, these may

可能存在的问题了
be problems

所以我们这样的话
In this case

方便抽样它当然就属于
convenience sampling, of course

非概率抽样
is non-probability sampling

就是它没有按照这个随机原则
It doesn't follow the random principle

去抽取样本
in sampling

这样的话它的这个样本的代表性
So, in terms of its sample representativeness

我们说就是有很大的这样一个问题的
we say there is a big problem

但是在现实生活当中
But in real life

因为为了操作方便
for convenient operation

还有一些成本的限制
and due to some cost restrictions

可能也会采取这种抽样方式
we may adopt this sampling method

再接下来是滚雪球抽样
Next, snowball sampling

滚雪球抽样的这个方法
Next, about snowball sampling

也是为了我们更好的去
In order to better

其实就更容易的去抽取样本单位
or be easier to get sample units

比如说我们先随机的
we may first randomly

选择一些被访者
select some respondents

然后对他实施访问
and interview them

再请他们再给我们提供一些
Then, ask them to recommend some people

他们认识的
that they know

他们比较熟悉的
they are familiar with

跟我们的研究目标一致的
who are consistent with our research targets

这样一些调查对象
as respondents

就提供给我们
According to their introduction

然后我们按照这个线索
we find some clues

再去调查这些单位
Then, we will go to investigate these units

这就是滚雪球抽样
This is snowball sampling

就是雪球越滚越大 越滚越大
The snowball will become bigger and bigger

比如说我要研究这个
Let's say I want to study

退休老人的生活
life of retirees

那么我可以清晨到公园
Then I can go to the park early in the morning

去结识几位散步的老人
get acquainted with some old people walking

那我调查完这几个老人之后
So when I finish an investigation on these old people

我可以让他们再给我推荐
I can ask them to recommend me

他们的朋友
their friends

那么可能不要很久
Then maybe not for long

你就会有一大批的老人朋友
you will have a large number of elderly friends

来帮助你做这个调查
to help you with the investigation

但是这样的调查
But what’s the problem

存在的问题是什么呢
with such investigation

就是我们要调查
We want to investigate

这个退休老人的这个生活
the life of retirees

那我们退休老人是不是
Then, will old retirees

一定会去公园
be sure to go to the park

肯定有一些老人他不喜欢活动
There must be some old people who don't like activity

不爱去公园
don't like going to the park

不爱和别人交往
don't like to socialize with others

那么就喜欢在家里
They like being at home

那么这样的话
In this case

你的样本就很难去代表
your samples are hard to represent

这部分的总体
the population of this part

那么这样的话
In this case

样本的代表性
the sample representativeness

也是存在一定问题的
still has certain problems

但是确实我们说
But, really

就比较方便展开这个抽样的
it's easier to expand this sampling

因为越滚越大 越滚越大
Because the more it rolls, the larger it will become

那么再接下来
Then, next

还有一种常见的
there is still another common

这个非概率抽样的方式
non-probability sampling method

就是这个定额抽样
That is quota sampling

那么定额抽样是
About the quota sampling

我们也是按照这个
we also, according to

调查对象的整体单位的特征
the characteristics of the overall unit of the investigation object

先将总体分成若干类
divide the population into several categories

然后再按一定的比例
Then, according to a certain proportion

在各个类别当中
in various categories

我们分配这个样本单位的数额
we distribute the quotas of this sampling unit

然后再去抽样
Then, we will do sampling

但是我这个抽样是主观抽样
However, this is a subjective sampling

那也就是说还是不遵循
That is to say, it still does not

随机的原则
follow the principle of randomness

比如说我们大家经常展开一些
We may often conduct

社会调查
social investigations

其实你在你这个调查问卷的最后
Actually, behind the questionnaire

你说我有一个配额
you say I have a quota

我怎么配额呢
How do I set the quota

我说男女比如说各占一半
I say it's a 50-50 split among men and women

然后收入我分了各个的这个阶段
Then, I divide the levels of income

比如说两千元以下
Let's say under two thousand yuan

两千元到三千元等等
two thousand to three thousand yuan and so on

然后我要在两千元以下
Then, from those with income of below two thousand yuan

比如说抽10%
I’ll choose 10%

两千元到三千元抽10%等等这样
choose 10% from those with income between two thousand and three thousand yuan, and so on

那么这个事实上都是你先做的
In fact, you do it in advance

这样一个配额
You set up a quota

你要按照这个比例去抽
You choose according to this proportion

但是这个定额抽样
However, this quota sampling

我们说跟分层抽样的区别就是
is different from stratified sampling in that

虽然它们都分了类
Though categorization is made in both methods

但是定额抽样
for the quota sampling

它在具体抽的过程当中
in the specific sampling

是你主观去控制的
you control it subjectively

就是它不是随机去抽的
It's just that it's not random

比如说我就是说男生女生
For example, you sample male students and female students

各占一半
in a half to half ratio

你自己主观去
You subjectively find

比如说找 找50%的男生
50% male students

找50%的女生
and 50% female students

就是说
That is to say

你不是完全按照随机原则来做的
you don't do it completely by random

所以这个就是定额抽样
So, this is quota sampling

下面我们再来看一个
Let’s see another sample of

定额抽样的例子
quota sampling

就是假设有一个高校
Let’s say there is a university

有三千名学生
that has a total of three thousand students

男生占了60%
Male students account for 60%

女生占了40%
Female students account for 40%

文科和理科各占50%
Students of arts and sciences each account for 50%

然后一年级占了40%
Freshman students account for 40%

二年级 三年级 四年级
Sophomore students, junior students and senior students

分别又占了30% 20%和10%
account for 30%, 20% and 10% respectively

那么这个时候
In this case

你如果用定额抽样的方法
if you use the quota sampling method

去抽取一个样本容量
to get a sample size

为100人的样本的话
of 100 students

那么我们就可以得到
Then, we can get

下面的定额表
the following quota table

其实大家看我们就是按照
In fact, we determine

完全按比例来确定这个样本单位
the sample units fully in accordance with the proportions

但是在操作的时候
However, in operation

我们是主观去选取
it is subjectively that we choose

这些样本单位的
these sample units

这跟分层随机抽样
Compared with the stratified random sampling

是有着明显的不同的
It
is evidently different

这个是需要大家注意的
This needs our attention

这就是我们要给大家讲的
These are some common sampling organization forms

一些常见的抽样组织形式
I want to introduce

[A1]随机
[A2]There are also some problems with the representativeness of the samples.