Statistics

• Census


  Observes every member of a population to give a completely accurate result, but is expensive, time-consuming and destroys the good
• Sample


  Observations of a subset of the population which is cheaper, quicker and easier, but may not be representative of small-sub groups
• Simple Random Sampling


  -Free of bias

  -Easy and cheap for small populations

  -Requires a sampling frame
• Systematic Sampling

  First item is chosen randomly then the rest are chosen at regular intervals of k = population/ sample size

  -Requires a randomised sampling frame to avoid bias

  -Easy to use for large populations
• Stratified Sampling


  Population is divided into mutually exclusive strata and a random sample is taken from each sample= no. in stratum/no. in population x sample size

  -Accurately reflects population and proportions

  -Requires a sampling frame and population must be categorised
• Quota Sampling


  Population is divided into groups and are observed until the quota is fulfilled

  -small sample can be representative

  -no sampling frame

  -quick, easy, cheap

  -bias due to non-random sampling

  -population must be divided into groups

  -larger population means more groups which increases costs
• Opportunity Sampling


  Observing those in the population who are available at a given time and fit the criteria

  -Easy and cheap

  -Unrepresentative

  -Interviewer bias
• Continuous Variable


  Can take any value in a given range
• Discrete Variable


  Can only take specific values in a given range