Sampling

• Define sampling:

  Selecting a group from target/survey pop. to study
• How should a sample be chosen?

  • To be representative of the target pop.
  • W/ a sampling frame
• What is a sampling frame?

  List of target pop.
• 3 examples of sampling frames:

  1. Electoral Register
  2. School/uni register
  3. Company payroll
• 3 problems w/ sampling frames?

  • Could be outdated
  • No lists for groups such as homeless
  • Access denies due to legal reasons, confidentiality etc.
• How would Positivists research?

  • Macro-level research
  • Large target pop.
  • Use representative sample for generalisable results
• How would Interpretivists research?

  • Micro-level research
  • Small target pop.
• Opportunity sampling

  • Taking advantage of situations to identify individuals who might participate in research (e.g. researchers interested in voting behaviour could target ppl leaving polling stations on a Gen. Election
• PROs + CONs of opportunity sampling:

  +VE: Easy + quick

  -VE: Unrepresentative + risk of bias (researcher may only approach non-threatening ppl)
• Purposive sampling

  • Purposefully choosing ppl w/ characteristics the research is focussed on (e.g. targeting druggies for study on drug use)
  • You could track them down to invite them
• Famous study using purposive sampling?

  Willis' 'Learning to Labour'
• Snowball sampling

  • Type of purposive sampling
  • Contacting 1 individual + asking them to recruit others that are suitable
  • Useful when no sampling frame
• Which perspective is most likely to use snowball sampling?

  Interpretivists
• Volunteer sampling

  • Advertising research to ask volunteers to participate
  • Common approach is to send invites via email
• 2 CONs of volunteer sampling?

  • Researcher can't control who gets recruited
  • Sample will unavoidably be biased (e.g. ppl are more likely to volunteer if it's paid, if they're interested in the topic etc.)
• Random sampling

  Selecting via varied methods from a complete sampling frame
• PRO of random sampling?

  Removes ALL bias from selection process
• 2 CONs of random sampling?

  • Requires complete sampling frame
  • Sample could be unrepresentative
• Stratified sampling

  • Dividing target pop. into 'strata'
  • Sample should have same strata in same proportions
  • Strata are then filled randomly
• 2 CONs of stratified sampling?

  • Requires complete sampling frame
  • Requires lots of info ab. target pop. that might be unavailable
• Systematic sampling

  • Choosing every Nth person from sampling frame
  • Same PROs + CONs as random sampling (removes bias BUT unrepresentative sample)
• Quota sampling

  • Filling strata that correspond to target pop. via non-random methods, e.g. purposive, volunteer, opportunity
• PRO + CON of quota sampling?

  +VE: Helpful when you need a representative sample BUT incomplete sampling frame

  -VE: Inevitable bias
• Who often uses quota sampling?

  Companies trying to gauge public op.

  • Approach ppl in street but get certain numbers from diff. demographics
• What is meant by access?

  Contacting potential respondents to recruit them into a sample.
• 3 groups that researchers will have issues w/ sampling:

  Protected groups - often protected for safeguarding e.g. children, prisoners, those w/ mental issues. A gatekeeper (e.g. parents, doctor) permits access.

  Hidden groups - hard to locate + contact e.g. deviants. Gatekeeper needed again to put researcher in touch.

  Hostile groups - don't want to be researched e.g. cults