sampling

• Sampling

  • Ensure the group being studied are representative of the population as a whole.
  • This allows findings to be generalised about the population as a whole.
  • When studying educational factors it may not be possible to interview every student in the school.
  • A representative sample is extremely useful; practical limitations!
  • Representative samples are preferred by positivists as the findings can be generalised.
• Sampling

  The sampling Frame

  • Everybody in the population makes up the research population.
  • From the research population those who are of interest of study are identified- the sampling frame.
  • From the sampling frame identity the sample.
  • From the sample responses received are from the respondent population.
  • Information to identify the sample frame can be obtained from the electoral register, census, or other official statistics, which should be the most up-to-date version available.
• Sampling

  Types of representative sampling:

  • Random
  • Systematic or Quasi
  • Stratified random
  • Quota
• Random

  • The most straightforward types of sampling where those from the sampling frame are selected at random by chance.
  • Names from a hat; Wheel of Names.
  • Everyone in the sample frame has an equal chance of being selected.
  • A random sample should be large enough to represent the research population.
• Systematic or Quasi

  • Every nth person in the list is selected.
  • Young and Wilmott (1962) every 36th name.
• Stratified random

  • The sampling frame is stratified by characteristics i.e. gender, age, social class, and then the sample is proportionally created to represent it.
• Quota

  • The sample frame is stratified as before, and the researcher is given a quota of X number of people to study who fit the sample frame.
  • The researcher finds their own respondents and continues until the quota is met.
  • Due to theoretical and practical issues it is not always possible to achieve a representative sample- non-representative sampling.
  • Samples which are non- representative are less valid than samples which are representative.
• Theoretical issues

  • A researcher may adopt a method which does not create a representative sample but uses their preferred method of data collection.
  • Qualitative or quantitative; positivists or anti-positivists.
• Practical issues

  • The characteristics or demographics of the research population may not be known.
  • Some sampling frames do not have an accessible registry from which a sample can be created i.e. criminals.
  • Respondents may refuse to participate.
• To overcome practical issues with non-representative sampling, there are two types of non-representative sampling:

  1. Opportunity.
  2. Snowball.
• Opportunity

  • Choosing the individuals who are easiest to access.
• Snowball= Asking respondents if they know others who would like to take part in the study, creating a snowball effect.