Sampling Design

Question 1

What makes a good representative sample?

Answer 1

• Large
• Unbiased
• Random

Question 2

What is selection bias and sampling error?

Answer 2

• In a random sample, sampling error is inevitable
  
  • Sampling error: difference between the value of a statistic and the parameter (due to chance)
• In a nonrandom sample, selection bias is inevitable
  
  • Selection bias: systematic favouritism in the data selection process leading to misleading results

Question 3

Define unit

Answer 3

Unit: an individual in a sample obtained from the sampling frame which receives the treatment

Question 4

Compare random sampling and simple random sampling

Answer 4

Random: each unit in the population has a chance of being selected

Simple random: each sample has an equal chance of being selected

• All SRS are random

Question 5

Compare stratified and cluster sampling

Answer 5

Both populations are divided into subgroups based on pre-existing characteristics

• Stratified: units are drawn from each strata
  
  • Not random unless proportionate sampling
• Cluster: a sample of subgroups are censused

Question 6

What is multistage sampling?

Answer 6

• Multistage: taking samples in stages using smaller and smaller sampling units at each stage

Question 7

What is systematic sampling?

Answer 7

• Systematic: a unit is selected as the starting point, then additional units are sampled at a constant interval, k
  
  • k = total population/sample size we want

Question 8

Compare response bias and nonreponse bias

Answer 8

• Nonresponse: a selected individual refuses or cannot be contacted (also called undercoverage)
• Response: influence of the behaviour of respondent or interviewer