The research question will suggest much about the methods required and the methods chosen must provide valid data for the subject studied.
Research methods need to be:
- Appropriate for the setting;
- Those that will answer the question;
- Achievable in relation to the resources available (often most significantly in relation to time and finance);
- Acceptable to participants, and fall within ethical codes for the field;
- Appropriate to the researcher's skills (but also suit their style of working).
In broad terms, research methodology may be:
- Quantitative: data is in the form of numbers, eg the number of times a child shows a particular behaviour;
- Qualitative: data is in the form of opinions and experiences, eg parent/carer perceptions of their son/daughter's transition process;
- A mixture of both.
It is not necessary to use only one method; in fact researchers commonly use a combination. An inquiry may explore an issue (for example, 'what are the most challenging times of day to engage pupils in sedentary activities?') through a questionnaire, and then seek to explain it through follow-up interviews and observations.
The purpose of a research study will shape the methodology used, depending on whether the inquiry seeks to:
- Explain (the causes/consequences/processes of something);
- Predict (an outcome);
- Explore (the experience of something);
- Evaluate (the efficacy/efficiency/worth of something);
- Describe (the nature of something, how something happens);
- Test (a theory/assumption).
The way a research question is posed will affect the methodology used to answer it. It is important to identify exactly what you are attempting to find out.
- What causes a cup of water to overflow? (exploratory);
- How much water causes a cup to overflow? (test);
- What happens when a cup of water overflows? (explanation);
- Is cup A more likely to overflow than cup B? (comparative evaluation).
All of the examples are studies of a cup of water overflowing. But the nature of the question suggests different methods to seek an answer.
Even within a particular approach, a number of factors can still potentially affect the results of a research study.
An exploratory study of pupil attitudes towards exams may yield different results depending on whether a questionnaire, individual interviews or focus groups are used.
Identify how you think that these different approaches may influence findings.
Interviews may provide more detailed explanations than even open-question questionnaires
since there is potential flexibility in the question phrasing. Some pupils will give
more information in a face-to-face interview than when expected to write answers out.
Conversely, some pupils will give more information via a questionnaire than when being
interviewed. Individuals who dislike exams may also be more reluctant to respond to
questionnaires. Some pupils may be more reserved speaking in a
A mixture of qualitative and quantitative methodology can have a number of advantages, including:
- If the results of both methods relate to the same issue, each can confirm the accuracy of the other (known as triangulation);
- Quantitative data may provide the outcome ('what'), which can then be supported by qualitative data to explain the process ('why' and 'how');
- Quantitative data often focuses on the researcher's concepts, which can be moderated by qualitative data giving others' perspectives;
- Qualitative studies often need to be small scale, but additional quantitative data can support generalisation of findings.
Research projects come in a multitude of types and sizes. When you read published
research, the researchers may describe their studies using descriptions such as experiment, quasi-experiment, case study, action research, ethnographic study or grounded theory, amongst others. What do these terms mean?
Whilst class-based or school-based inquiry tends to be of small scale, there is still a range of approaches that could be applicable in these settings.
Find out more about this area and mixing and matching research methods in the following document: Research designs.