Samples and Populations

• Populations


  The entire collection of people, animals, plants or objects that we are interested in, sharing a common characteristic

  Defined by population parameters (measurements which describes the pop.)

  Vary in size- all UoM students vs all UoM students studying Psych
• Samples


  A selection of individuals from the larger pop., for any pop. there are many possible samples, will vary in size

  Defined by sample statistics (measurements which describe the sample, used to infer pop. parameters)

  Must be representative and free from bias
• Why do we sample?

  Time- difficult to collect data from everyone

  Money- expensive to collect data from everyone

  Access- not always possible to reach all members of a pop.

  Sufficiency- patterns of results don't always change much even if there is data from everyone
• How do we sample?

  Random

  Systematic

  Stratified

  Cluster

  Opportunity/Convenience

  Snowball
• Random sampling

  The gold standard, each member of pop. has an equal chance of being selected, usually quasi-random
• Systematic sampling

  Draw from the population at fixed intervals, problematic in populations with a periodic function
• Stratified sampling (when we know something)

  Proportional: specified groups appears in numbers proportional to their size in the population

  Disproportional: specified groups which are not equally represented in the population, are selected in equal proportions
• Cluster Sampling


  Researcher samples an entire group or cluster from the population of interest
• Opportunity/Convenience sample


  People who are easily available e.g. all who attended a live lecture, but can lead to a biased sample
• Snowball sampling


  Recruit small numbers of Ps and then use those initial contacts to recruit further participants

  Biases the sample, but useful if you want to recruit very specific pop.
• External validity 


  Gather data from samples in order to infer pop. parameters

  External validity: ability to generalise results

  Population validity: Is our sample representative

  Ecological validity: does the behaviour measured reflect naturally occurring events
• Internal vs External Validity 


  Experimental designs= high internal but low external validity

  Naturalistic designs= low internal but high external validity
• Sample size


  Size does matter

  The smaller the sample the bigger the sampling error

  Battle between size vs time/cost

  Factors in deciding sample size:

  Design- subjects design, number of IV's or IV levels

  Response rate

  Heterogeneity of pop.- Diverse population needs a large sample in order for it be representative