Linking the research question to study design
Making the research goals or aims operational is the next step. One of the essential parts of planning the project is to get its size right – you need to know that the research is feasible and that it can be done within the resources you have available. This means getting the level of the question right and having a really clear focus. When we start thinking of what we’d like to do, nine times out of ten we’ll be thinking too big. Getting the right level for research usually means talking the idea through with others, writing out your ideas as a brief proposal with clear aims, and getting more formal feedback. The more time that is put in now, the less will be spent feeling despondent or overwhelmed in future.
The size of the project – that is, the scope defined by the research question asked – will depend partly on:
- time available
- the availability of internal or external funding and support (e.g. grant money)
- whether this should be a pilot project or build on other things
- your research ‘sample’ and access to participants
- whether the topic calls for cross-sectional or longitudinal designs and how this affects your availability and needs
- whether this is to meet a higher degree or not.
- The simplest level of assessment would be in terms of judging the participants’ immediate satisfaction with the course.
- The next level might be to address whether a change in knowledge has taken place.
- The next might involve considering whether a change in stress levels has occurred, after the course and six months later.
- The next level (an ambitious one) might look for relationships between changes in a specific patient-related outcome (e.g. complaints, patient satisfaction, quality of life, medication error) and changes in either or both knowledge and stress levels of the participants.
- The largest question would be to ask if the course has produced long-term effects on participants and patients. This most ambitious level would involve conducting a trial using the same outcomes as above, but randomly selecting those in the intervention group and those who become controls. Your choice of level would depend on the factors outlined above (Jones et al., 1988).
As you see, each research question implies a different design and a different magnitude of commitment. Bordage and Dawson (2003) suggest that we ask the following questions about research design at this stage.
- Do you want to intervene or simply observe?
- Do you need a control group?
- How will you control for confounding variables?
- What is the ‘best’ research design to answer your question?
And the following questions about the sample of research participants.
- What are your criteria for inclusion and exclusion of participants?
- How are you going to obtain participants?
- If an experiment, how will you assign participants?
- How many participants do you need? How will you address issues relating to power and researcher bias?
Being able to answer these questions will help frame the research approach, study, methods and timeframe.