Sampling

Survey population

• All members of the population one is interested in studying (e.g. all W/C boys in the UK)

Sample unit

• Individual w/in a sample

Sample attrition

• Sample unit leaving

Representative sample

• Reflects the composition of the group being studied

4 representative sampling methods

1. Random sampling

2. Systematic/quasi-random sampling

3. Stratified sampling

4. Quota sampling

Random sampling - description, advantages and disadvantages

• Sample drawn by chance (e.g. out of a hat)

• Advantages:

  • No selection bias

  • If large enough, sample will be representative (able to generalise)

• Disadvantages:

  • Time consuming

  • Requires a complete list of the population

  • If not large enough, sample won’t be representative (unable to generalise)

Systematic sampling - description, advantages and disadvantages

• Every nth person selected

  • Wilmott and Young - 36th

• Advantages:

  • Simple

  • Ensures a spread across the population

• Disadvantages:

  • Risk of periodicity

  • Not fully random

  • Requires a complete list of the population

Stratified sampling - description, advantages and disadvantages

• Survey population stratified into age or class or gender

  • The composition is noted, and the sample made to reflect this

    • E.g. 20% of the pop is under the age of 18, so 20% of the sample is

• Advantages:

  • Very representative of subgroups

  • More precise

• Disadvantages:

  • Requires detailed pop info

  • Way more complex

Quota sampling - description, advantages and disadvantages

• Population stratified into age or class or gender etc.

  • Interviewer given a quota of e.g. 20 men and 20 women and have to choose respondents that fit these characteristics

    • Continue until all quotas are full

• Advantages:

  • Quock and cost-effective

  • Ensures representation of specific subgroups

• Disadvantages:

  • Not truly random sampling due to subjective selection

  • Interviewer may just choose the closest people, so won’t be truly representative (time or geographical bias)

Interpretivist view on representative sampling

• :|

  • Don’t really care because have no interest in making generalisations

Positivist view on representative sampling

• :)

  • Allows them to make accurate generalisations

3 practical issues with (representative) sampling

1. Social characteristics of ther research population may nor be known

2. Can be difficult to find/create a sampling frame

• E.g. not all criminals are convicted

3. Potential respondents may not want to participate

• E.g. criminals

Un-representative sampling

1. Snowball sampling

2. Opportunity sampling

Snowball sampling - description, advantages and disadvantages

• Choosing key individuals who nominate others to be interviewed

• Advantages:

  • Cost effective

  • Useful way to contact those difficult to find/persuade

• Disadvantages:

  • Unrepresentative as the people who know each other will share characterstics/values (ungeneralisable)

  • More biased 

Opportunity sampling - description, advantages and disadvantages

• Choosing individuals who are easy to access like passerbys

• Advantages:

  • Easy

  • Quick

  • Cheap

• Disadvantages:

  • Unrepresentative

  • Biased