Tuesday, July 17, 2012

Sampling(statistics)

Did you remember the question we got when we did the body measurement last time?
The question is like "Is this sampling random? Concerning about the gathering of the elemental selection."

I found out there are four types of random sampling:

1. simple random sampling: all elements in the frame have an equal probability of selection. 
    ->not this one

2. systematic sampling: selecting the elements according to a planed scheme, like n+5, n+10, n+15......
    ->not this one either

3. stratified sampling: if the population has a number of distinct categories, then the population can be separated by these categories into different "strata." then these stratum are considered to be some independent sub-population, out of which individual elements can be randomly selected. then, we just  do the simple random sampling in each strata. 
    -> ahha, I think is this one!

4. cluster sampling: cluster sampling is sort of the same with stratified sampling; the different part is that the elements in each cluster must be heterogeneous. that way, each cluster can be taken as a whole population, and all we have to do is a simple random sampling in only one cluster.
    -> not this one

therefore, in this case, they can do the stratified sampling separating Taiwan into some geographical regions, like north, middle, and south. and do simple random sampling in each.(幹太累了我不想打英文了)不過我覺得還是會不準,因為他只有做北中南,缺了東部。然後還有一個主要的問題,集中做都市的樣本不是random selection阿,除非他把北中南三區內再區分各clusters才可以吧.................


2 comments:

  1. 重點就是,他們根本不是隨機挑人阿,這樣根本沒辦法代表母體阿,如果是隨機挑選的話,他們要分成那些年齡層,就上述來講,應該是沒問題的。重點就是,沒有隨機。

    ReplyDelete
  2. 我今天再和我數學系的同梯討論了一次。
    對,沒有隨機。

    ReplyDelete