LICENSE

Reasons of Taking a Sample:

• Destructive (eg: cannot eat all chocolates in a factory because you will die)
• Time Consuming
• Expensive
• Condition Change (eg: environmental)
• Not all population available

Sampling Methods: (in pg11~12 of inference book)
Must justify the reasons of choosing your sampling method

• Simple Random (the most ideal)
  • ‘Random’ selection: Assign every member of the population an unique ID (PK),
    generate randint (ignore repetition) until sample size reached.
    “Every member of the population has an equal chance of being included.”
• Cluster
  • Taking a cluster out of the population
• Systematic
  • eg: every 3rd person
• Stratified (ideal under some circumstances)
  • Balance eg: every 2 boys selected, one girl selected (contingent that there are more girls than boys with a ratio of approximately 2:1)
  • Calculation Required

Cluster & Systematic sampling methods is good, but has a high chance of creating bias.

Simple and Stratified sampling methods are the most ideal as they give us an unbiased sample.
Although simple random sampling methods are the best, stratified sampling methods are better if another variable (eg: age)
can create a significant difference in the sample. Example of this could be:

However, I could’ve improved this sample by using the stratified sampling method, 
since the people’s gender could have significantly affected this sample since 
most people tend to save the young, old, then the woman when it comes to dangerous 
situations. Therefore, using the stratified sampling method would improve this 
simple sampling method. 

Author: 영
Source: wndnotes
Link: https://discord.gg/yHdQrahnAJ

more info on stratified sampling can be found in the next post: