Selection

Question 1

Sampling

Answer 1

Selection of a representative part of a population to test data, then generalise results

Question 2

Sampling frame

Answer 2

Individuals in a population that are eligable to be included in a sample

Question 3

Non-probability sampling & examples

3

Answer 3

Individuals are selective based on subjective methods (limited representation but good when some individuals have 0 chance of being picked or unknown parameters)

Convenience: chosen based on convenience (over and under representation, self selection bias)
Purposeful: chosen by someone familiar with the population (limited generalisability)
Snowball: start with one person who recommends others

Question 4

Probability sampling & examples

3

Answer 4

Random selection where each unit has equal probability of selection
Simple random: choice based on chance
Systematic random: random choice then selection at regular intervals
Stratified: simple random from each population strata (gender, age)
Cluster: random sampling from each heterogenous group (school, hospital)

Question 5

Sampling error, types and how to mitigate

2

Answer 5

• errors in sampling that leave a non-representative population
• Chance: random sampling errors that are mitigated by using a large sample
• Bias: subjective influence due to poor sampling method

Large sample size, defined ex/inc criteria, power analysis, strong sampling technique

Question 6

Non-sampling error

Answer 6

observation error (uncalibrated equipment, interviewers effect, respondant effect)