Sampling methods

• Simple random sampling: Every person/item in the population has an equal chance of being in the sample, and each selection is independent of the others.
• Simple random sampling - method:

  • Give a number to each population member
  • Generate a list of random numbers and match them to the numbered members
• Simple random sampling:

  • Advantage - unbiased - every member has equal chance
  • Disadvantage - can be inconvenient if population is spread over a large area
• Systematic sampling: Selecting every nth member from a population
• Systematic sampling - method:

  • Number each member of the population
  • Calculate a regular interval by dividing the population size by the sample size
  • Generate a random start point to choose the 1st member
  • Keep adding the interval to select the sample
• Systematic sampling:

  • Advantage: can be set up by a machine for quality control
  • Disadvantage: if interval coincides with a pattern, sample could be biased
• Stratified sampling: If population is in categories, use the same proportion of each category in the sample as in the population.
• Stratified sampling - method:

  • Divide the population into categories
  • Calculate number needed for each category: (size of category in population/population) x sample size
  • Randomly select the sample for each category
• Stratified sampling:

  • Advantage: if categories are disjoint (no overlap), sample is representative
  • Disadvantage: extra detail needed can make it expensive
• Quota sampling: Used in market research - people are interviewed until a quota for each category is filled.
• Quota sampling - method:

  • Divide population into categories
  • Give each category a quota
  • Collect data until the quotas are met in all categories
• Quota sampling:

  • Advantage: easy - dont need access to the whole population
  • Disadvantage: can be biased as selection isnt random
• Opportunity sampling: Sample is chosen from a selection of the population that is most convenient.
• Opportunity sampling - method:

  • Choose the members of the population that are the easiest to sample
• Opportunity sampling:

  • Advantage: data can be gathered quickly and easily
  • Disadvantage: is not random so can be very biased - no attempt to make the sample representative
• Cluster sampling: Population divided into distinct groups - clusters are groups expected to give similar results.
• Cluster sampling - method:

  • Divide whole population into clusters
  • Randomly select clusters for the sample, based on required size
  • Either use all members from selected clusters or randomly sample within each cluster
• Cluster sampling:

  • Advantage: more practical - can incorporate other methods to make it adaptable
  • Disadvantage: only sample certain clusters so less representative
• Self-selection sampling: (Volunteer sampling) People choose to be part of the sample.
• Self-selection sampling - method:

  • Advertise to whole population for participation
  • Either use everyone who responds or take a sample to best represent the population
• Self-selection sampling:

  • Advantage: requires little time or effort
  • Disadvantage: can easily have trends, such as strong opinions which leads to bias