How To Choose Your Participants
Dr Catherine Dawson has worked as a researcher since the mid-1980s and has taught on research methods courses at university. She has also written extensively for academic journals on a wide range of subjects including research methodology. She is based in Weymouth, Dorset.
As you continue planning your research project you need to think about how you’re going to choose your participants. By now you should have decided what type of people you need to contact. For some research projects, there will be only a small number of people within your research population, in which case it might be possible to contact everyone. This is called a census. However, for most projects, unless you have a huge budget, limitless timescale and large team of interviewers, it will be difficult to speak to every person within your research population.
UNDERSTANDING SAMPLING TECHNIQUES
Researchers overcome this problem by choosing a smaller, more manageable number of people to take part in their research. This is called sampling. In quantitative research, it is believed that if this sample is chosen carefully using the correct procedure, it is then possible to generalise the results to the whole of the research population.
For many qualitative researchers however, the ability to generalise their work to the whole research population is not the goal. Instead, they might seek to describe or explain what is happening within a smaller group of people. This, they believe, might provide insights into the behaviour of the wider research population, but they accept that everyone is different and that if the research were to be conducted with another group of people the results might not be the same.
Sampling procedures are used everyday. Market researchers use them to find out what the general population think about a new product or new advertisement. When they report that 87% of the population like the smell of a new brand of washing powder, they haven’t spoken to the whole population, but instead have contacted only a sample of people which they believe are able to represent the whole population.
When we hear that 42% of the population intend to vote Labour at the next General Election, only a sample of people have been asked about their voting intentions. If the sample has not been chosen very carefully, the results of such surveys can be misleading. Imagine how misleading the results of a ‘national’ survey on voting habits would be if the interviews were conducted only in the leafy suburbs of an English southern city.
Probability samples and purposive samples
There are many different ways to choose a sample, and the method used will depend upon the area of research, research methodology and preference of the researcher. Basically there are two main types of sample:
- probability samples
- purposive samples.
In probability samples, all people within the research population have a specifiable chance of being selected.
TABLE 3: SAMPLING TECHNIQUES
PROBABILITY SAMPLES |
PURPOSIVE SAMPLES |
The researcher is interested in finding out about national detention rates. He wants to make sure that every school in the country has an equal chance of being chosen because he hopes to be able to make generalisations from his findings. He decides to use a simple random sample. Using this method the researcher needs to obtain the name of every school in the country. Numbers are assigned to each name and a random sample generated by computer. He then sends a questionnaire to each of the selected schools. The researcher would have to make sure that he obtained the name of every school in the country for this method to work properly. |
The researcher decides that he wants to interview a sample of all pupils within a school, regardless of whether they have been on detention or not. He decides to use a quota sample to make sure that all groups within the school are represented. He decides to interview a specified number of female and male school pupils, a specified number of arts, sciences and social science pupils and a specified number within different age categories. He continues approaching students and interviewing them until his quota is complete. By using this method only those pupils present at the same time and in the same place as the researcher have a chance of being selected. |
The researcher wants to find out about national detention rates, but is interested also in finding out about school policy concerning detention. He decides that to do this he needs to visit each selected school. To cut down on travel costs, he decides to use a cluster sample. Using this method, geographical ‘clusters’ are chosen and a random sample of schools from each cluster is generated using random number tables found at the back of some statistics books. Using this method the researcher only needs to travel to schools within the selected geographical regions. The researcher would have to make sure that he chose his clusters very carefully, especially as policy concerning detention might vary between regions. |
The researcher is interested in carrying out semi-structured interviews with pupils who have been on detention over the past year. However, he finds that the school has not kept accurate records of these pupils. Also, he doesn’t want to approach the school because he will be seen by the pupils as an authority figure attached to the school. He decides that a snowball sample would be the most appropriate method. He happens to know a pupil who has been on detention recently and so speaks to her, asking for names of other pupils who might be willing to talk to him. The researcher should obtain permission and have a chaperone or guardian present at the interviews. He needs to be aware also that friends tend to recommend friends, which could lead to sampling bias. |
The researcher has decided that he wishes to conduct a structured interview with all the children who have been on detention within a year at one school. With the head teacher’s permission, he obtains a list of all these pupils. He decides to use a quasi-random sample or systematic sample. Using this method he chooses a random point on the list and then every third pupil is selected. The problem with this method is that it depends upon how the list has been organised. If, for example, the list has been organised alphabetically, the researcher needs to be aware that some cultures and nationalities may have family names which start with the same letters. This means that these children would be grouped together in the list and may, therefore, be underrepresented in the sample. |
The researcher has heard of a local school which has very few detentions, despite that school having a detention policy. He decides to find out why and visits the school to speak to the head teacher. Many interesting points arise from the interview and the researcher decides to use a theoretical sampling technique. Using this method the emerging theory helps the researcher to choose the sample. For example, he might decide to visit a school that has a high detention rate and a school that has no detention policy, all of which will help to explain differing detention rates and attitudes towards them. Within this sampling procedure, he might choose to sample extreme cases which help to explain something, or he might choose heterogeneous samples where there is a deliberate strategy to select people who are alike in some relevant detail. Again the researcher has to be aware of sampling bias. |
The researcher has decided that he wishes to concentrate on the detention rates of pupils by GCSE subject choice and so decides upon a stratified random sample. Using this method the researcher stratifies his sample by subject area and then chooses a random sample of pupils from each subject area. However, if he found that there were many more pupils in the arts than the sciences, he could decide to choose a disproportionate stratified sample and increase the sample size of the science pupils to make sure that his data are meaningful. The researcher would have to plan this sample very carefully and would need accurate records of subjects and pupils. |
The researcher is a teacher himself and decides to interview colleagues, as he has limited time and resources available to him. This is a convenience sample. Also, at a conference he unexpectedly gets to interview other teachers. This might be termed haphazard or accidental sampling. The ability to generalise from this type of sample is not the goal, and, as with other sampling procedures, the researcher has to be aware of bias which could enter the process. However, the insider status of the teacher may help him to obtain information or access which might not be available to other researchers. |
TABLE 4: SAMPLING DOS AND DON’TS
DO |
DON’T |
Take time and effort to work out your sample correctly if you’re conducting a large scale survey. Read the relevant literature suggested in this book. Time taken at the beginning will save much wasted time later. |
Rush into your work without thinking very carefully about sampling issues. If you get it wrong it could invalidate your whole research. |
Discuss your proposed sampling procedure and size with your tutor, boss or other researchers. |
Ignore advice from those who know what they’re talking about. |
Be realistic about the size of sample possible on your budget and within your time scale. |
Take on more than you can cope with. A badly worked out, large sample may not produce as much useful data as a well-worked out, small sample. |
Be open and up front about your sample. What are your concerns? Could anything have been done differently? How might you improve upon your methods? |
Make claims which cannot be justified nor generalised to the whole population. |
Use a combination of sampling procedures if it is appropriate for your work. |
Stick rigorously to a sampling technique that is not working. Admit your mistakes, learn by them and change to something more appropriate. |
These types of sample are used if the researcher wishes to explain, predict or generalise to the whole research population. On the other hand, purposive samples are used if description rather than generalisation is the goal. In this type of sample it is not possible to specify the possibility of one person being included in the sample. Within the probability and purposive categories there are several different sampling methods.
The best way to illustrate these sampling methods is to take one issue and show how the focus of the research and the methodology leads to the use of different sampling methods. The area of research is ‘school detention’ and in Table 3 you can see that the focus and sampling techniques within this topic can be very different, depending on the preferences of the researcher, the purpose of the research and the available resources.
CHOOSING YOUR SAMPLE SIZE
The first question new researchers tend to ask is ‘how many people should I speak to?’ This obviously depends on the type of research. For large scale, quantitative surveys you will need to contact many more people than you would for a small, qualitative piece of research. The sample size will also depend on what you want to do with your results. If you intend to produce large amounts of cross tabulations, the more people you contact the better.
It tends to be a general rule in quantitative research that the larger the sample the more accurate your results. However, you have to remember that you are probably restricted by time and money – you have to make sure that you construct a sample which will be manageable. Also, you have to account for non-response and you may need to choose a higher proportion of your research population as your sample to overcome this problem. If you’re interested in large-scale quantitative research, statistical methods can be used to choose the size of sample required for a given level of accuracy and the ability to make generalisations. These methods and procedures are described in the statistics books listed at the end of this chapter.
If your research requires the use of purposive sampling techniques, it may be difficult to specify at the beginning of your research how many people you intend to contact. Instead you continue using your chosen procedure such as snowballing or theoretical sampling until a ‘saturation point’ is reached. This was a term used by Glaser and Strauss (1967) to describe that time of your research when you really do think that everything is complete and that you’re not obtaining any new information by continuing. In your written report you can then describe your sampling procedure, including a description of how many people were contacted.
SUMMARY
- If it is not possible to contact everyone in the research population, researchers select a number of people to contact. This is called sampling.
- There are two main types of sampling category – probability samples and purposive samples.
- In probability samples, all people within the research population have a specifiable chance of being selected. Only within random samples do participants have an equal chance of being selected.
- Purposive samples are used if generalisation is not the goal.
- The size of sample will depend upon the type and purpose of the research.
- Sample sizes should take into account issues of non-response.
- Remember that with postal surveys it might be difficult to control and know who has filled in a questionnaire. Will this affect your sample?
- In some purposive samples it is difficult to specify at the beginning of the research how many people will be contacted.
- It is possible to use a mixture of sampling techniques within one project which may help to overcome some of the disadvantages found within different procedures.
