When looking into primary research methods, it is useful to explore quantitative research. One good way of doing this is to use sampling. This means using a sample of the population to conduct your research. Let’s imagine that you want to explore whether school uniforms have a beneficial effect or detrimental effect on children and their parents. If you were using quantitative research, you may be looking at something that could be measured factually e.g. whether buying a uniform affects the family budget. To realise reliable results, you would have to ensure that your sample was big enough to be comprehensive but also little enough to be controllable. Added to that, you would also have to make sure that you covered a typical cross section of families.
Sampling can be divided in two groups – random sampling and non-random sampling. Random sampling can work very well, however, there are situations that need to be taken into consideration. Let’s imagine that you were out doing your research on the cost of school uniforms and you decide to hand out questionnaires at the school gates so that you can get large numbers. If you visit a school in an area that is struggling for employment, then the research will be invalid because you are not getting a cross section of society. Likewise, if you visited an expensive, private school, the chances are that you would be interviewing people that had similar financial backgrounds. All of this needs to be thought about before you embark upon your research.
Non Random sampling can be useful but it is difficult to know whether you are getting a general overview of the population. The reason for this is that you would use a list such as an election roll and choose every fifth name on the list. Can you see how this could be problematic because not every person that you will approach will have children that are in school? Obviously, if you could access a list of scout clubs or brownie groups that attracts children from different schools that would have more potential research value.
With both random sampling and non-random sampling, once you have got your sample of folks to research, you can either use questionnaires or interviews as mentioned earlier or you can test them. Testing would probably not work with the cost of the school uniform. However, if you were conducting how the effect of the school uniform worked on the psychological effects of the child, you might be able to use it. For instance, “do children respond better to tests when wearing the uniform or not?”
It is important to remember, when conducting your research and then writing it up that you must demonstrate that you’ve thought carefully about your research choices and why you have made them. You must also make it clear if you’ve recognised any weakness in your research and how it could be improved upon. This will demonstrate that you understand the research process and will help your marks.
Using interviews can be one the most productive and useful primary research methods. Interviews fall into three categories – structured, unstructured and semi structured.
- Structured interviews. (This is what you would use if you decide to use the sampling method which we discussed earlier.)
- All the interviewees would be asked the same questions and given a choice of answers.
- These interviews can be done by handing out paper forms, telephone calls or over the internet.
- The idea is to get a large numbers of responses so that you can quantify the data.
- This method has its flaws because often only a very small proportion of the data can be collected because the interviewees don’t bother to answer the questions. The other problem is that often the questions are misinterpreted and the researcher has no way of understanding this. Another flaw in this system is that because the interviewee might only have a choice of four answers, it might not actually fit their experience and so they give the wrong reply so that they can give an answer.
- A good insurance policy to take out for conducting structured interviews is to run a smaller pilot scheme first. This way you can find out what does and does not work. This may seem to be more work and more of your time being taken up but it can save both time, effort and heart ache overall.
- Unstructured interviews. This is the direct opposite of the structured interview. Whereas structured interviews collect quantitative data, unstructured interviews assemble qualitative data.
- The questions do not have closed choices but are open ended so that the interviewee can offer their opinions and feelings about the subject that you are researching.
- It is obvious in these situations if the subject does not understand the question or is withholding information because of the use of body language. This is one of the main advantages of unstructured interviews. You can access more clarity over the subject’s views.
- Questions needs to be designed very carefully to ensure that they are not leading the subject to please you or agree with you.
- A great deal of time can be lost if the subject goes off on a tangent and the topic that you are researching is forgotten. It is wise, therefore, to have questions ready that will bring the interviewee back to what is being discussed.
- When interpreting the interview, it is important that you don’t misunderstand the subject because of the regional accent.
- Interviews need to be recorded or filmed. Unlike the structured interview that has a small selection of answers, there is a lot more data to be worked through and if it isn’t recorded, it will be lost.
- Semi-structured interview – for some researchers, this is the route to go along because it is a mixture of tight questions and open questions.
- It is a way to gain quantitative data.
- It is also means that the more open questions might give you new ideas about your research.
- Designing the questionnaire can take more time and skill because it is useful for it to be balanced.
