Random Sampling

It is practically impossible to get a truly random sample. In a random sample every member of your target population would have an equal chance of being selected. So for example if you wanted a random sample of students from GIS you would need to obtain all of their names, put them in a hat and draw your sample out. In actual fact that would be the easy bit. The difficult task would be finding them and persuading their parents to let your chosen ones take part!

 

  • There is no bias in the way in which participants are chosen – everybody has an equal chance and no one is systematically excluded from the sample. Therefore, the sample is theoretically likely to be representative
  • It is clear to everyone how the sample was chosen – it can be easily explained and understood
  • There can be difficulty in obtaining the names of everyone in the target population, which may cause bias
  • Bias can arise if certain participants are chosen but cannot participate, for example if they are busy on the day in which the study is taking place, or simply if they don’t want to
  • The sample chosen by random sampling may not be useful, for example, if the study is investigating obedience within males and females between the ages of 20 and 40, random sampling may produce all male participants
  • Opportunity Sampling

    Opportunity sampling is less a mathematical method of sampling, and more of a “pick whoever is available” approach. Researchers will use whoever they can find who is filling to take part in the study. The ways in which the participants are chosen are not structured. An example might be someone doing questionnaires in a town high street. They will probably not have a specific participant in mind, but instead will just attempt to use everyone, and will happily include anyone who agrees to take part in the findings.

    • Can be ethical, for example, if the researcher is able to judge if the experiment will upset the potential participant
      or if they can work out if they will be too busy to participate in the research
    • The researcher has a lot of control over who is used to participate, and access to potential participants is not
      limited
  • Your participants may have something in common, as they were all in the same place at the same time. As a result your sample will be biased.
  • There is a lot of potential bias from the researcher – they may only choose people who are similar to themselves in some way, whether it be preferring people from the same sex or people of the same age as them
  • Volunteer (or self-selected sampling)

    Volunteer sampling calls for volunteers to willingly contribute their time for the study. They may respond to a letter inviting a number of people to participate, or more commonly might respond to an advertisement, which often involves payment also.

    • The most ethical form of sampling because the participants come to the researcher rather than the researcherseeking them out
    • Volunteers have self-selected themselves and are therefore most likely interested in the piece of research, this means that they will be less likely to give biased information or go against the researcher’s instructions
    • The volunteers are willing to be involved in the study
  • Can take a long time to get enough willing participants because the researcher has to wait for a response to their
    advertisement or letter
  • Because the participants have all selected themselves, they might all be similar in some way which will not provide a broad spectrum of results which is applicable to many other groups
  • Your participants were all motivated to take part, thus they already have something in common. As a result your sample will be biased. They may also perform better than the rest of the population due to being more motivated in general.
  • Stratified Sampling

    Stratified Sampling is used to ensure that certain groups are all represented by the sample. The researcher will decide what specific groups need to be tested within the sample, and will calculate how many people should be selected from each group using proportions. If you are investigating obedience in males and females between the ages of 20 and 40, you may separate them into four categories (males 20-29, males 30-40, females 20-29, females 30-40). If there were four times as many men as women, you would sample four times as many males.

     

    • Each group has to be represented by the sample, and so clear conclusions can be drawn outlining differences between the groups
    • Stratified sampling ensures that the right number of people from each group is chosen to represent their group. With random sampling, people from each group may still be picked but not in the correct proportions
  • It can be difficult to know how many to choose from each group to make the findings generalisable
  • The groups chosen by the researcher may not necessarily all be the important groups. Having the groups already decided means that some people will automatically be ruled out as participants