Research Guide – Framing a good research question.

by Dr Julian Quinn, The SERT Institute, RNSH.

All projects have at their heart a Question that they are designed to answer.

Asking the right question is, after all, the first step to solving a problem, but it goes further than that: a good research Question clarifies a project, and helps focus the thinking and planning. Constructing a research Question sounds straightforward but is usually not simple at all and, sadly, proper deliberation of the research Question is often overlooked. This can have important consequences, because if the Question is not clear and well thought through the whole project may be on shaky ground. It should be noted as well that the research Question has to be included in some form in project reports, publications and research ethics applications, so it must be clear and concise.

Another consideration at the outset is the source of the research Question – did a mentor or supervisor assign it? If so, it should not be taken for granted that it is cogent and well designed, although the chances are that it is. The mentor should discuss it in some depth, and perhaps some experienced colleagues should be consulted about it.

A research Question needs precision (but not too much)

It is very useful to be able to state the research question succinctly in a simple sentence. There can be exceptions but if it cannot be formulated in this way then it probably needs more thought. It must sound good and make logical sense. One type of phrase to avoid would be “is important in”, which is too imprecise to help define a goal. Thus “is factor X important in process Y?” is no good. Better is something like “does factor X have a significant impact on procedure Y outcomes under condition Z?” or “among patient group X, does factor Y show a strong correlation with feature Z?”

A research Question should not be too long

As a rough rule, it should not stretch to two full lines of standard typed text in length. However, it may have to be longer if a short preamble is needed to provide context. Or it may involve diseases with very long descriptors. The question can be split into parts and if necessary can refer to other projects, for example. Just be sure it is appropriate and makes sense.

A research Question should not be too vague or overarching

Research projects need a reasonable timeframe and realistic expectation of resource access. Big, bold, vague concepts have no place. Feasibility is crucial.

A research Question is usually quite different to a seminar title

Seminar titles tend to be bold and simply phrased to indicate the research area to be discussed even if that is broader than the subject of the seminar. It serves a different function to the research Question.

A research question should not be too parochial or small

Research conclusions by their nature have to extrapolate general or big things from smaller or more restricted things. That said, a research Question that is too obscure is unlikely to interest anyone.

Finding causes – do not to be tempted

There are many types of causes, and most diseases (and disease processes) have many causes and risk factors. Do not be tempted to state or imply that you are looking to define the cause of something. If you have a strong idea as to a putative cause, frame the question directly around that idea instead.

Questions point to the research aims

Note that stating a research Question in the preamble of a report or funding application will naturally flow into a statement of the research aims or objectives. One question may spawn several aims, which is great as long as it all makes sense and the connections are clear.

And lastly –

Try not to be boring. If the research question is answered, will anyone care? It can be a brutal question but it can be a very useful one. Be the one to ask it, not the audience at your final presentation.

Creative Commons License

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Leave a Reply

Your email address will not be published.