Resources > "HOW TO" Conduct Research > 2. Devise a research question

2. Devise a research question

Research questions usually emerge from critical thinking about your experiences in clinical practice. You may have noticed that a particular patient group responds in an unexpected way to a certain treatment, or you may have a ‘hunch’ that there’s a better way of doing something, or you may just have a strong interest in an area that you would like to investigate further. By conducting high quality research into the area you may be able to help not just your patients, but patients throughout the world.

(i) Review the literature
Research starts with a review of the literature. Enter your terms of interest into an electronic database (see below) to find out what’s been published in your area. There is no point re-inventing the wheel. If something has already been published, it’s unlikely that a repeat of the same research project will get funding or published.
PubMed -Searches MEDLINE and other life science journals for biomedical articles back to the 1950s
The Cochrane Collaboration - A collection of evidence-based medicine databases
Embase - Major bibliographic database for biomedical sciences
PsycINFO - Major bibliographic database for psychology
Also, attend conferences to learn about current research in your area of interest; who’s doing what, where it’s being done, and what they’re finding. This may help you to devise your own research question and provide you with a network of contacts with an interest in the same area as you.

Discuss your ideas with others to get their input as to whether or not it’s a FINER research question.[2]
Feasible. Is there adequate technical expertise, resources, funding, time and money available?
Interesting. Will the study be of interest to others beyond your practice?
Novel. Will the study provide new findings or will it confirm, refute or extend previous findings?
Ethical. Is the study likely to receive approval from an institution’s ethics board?
Relevant. Will the findings advance scientific knowledge, influence practice guidelines or guide future research?. For all research questions think about the implications of both positive and null (ie. no effect) findings. How will this knowledge contribute to, for example; practice, policy and future research?

(ii) Turn your ideas into a research question
When devising your research question, ask yourself the following:

  • What is your aim? (In general terms)
  • Are you looking to test a hypothesis? If yes, clearly state your hypothesis in the form of an answerable question or series of answerable questions
  • Are you looking to explore a new area to get an understanding of something we currently have little knowledge of?
  • Are you looking at a well researched area but would like to examine it in a new or novel way?
  • Are you looking to explore the relationship between two phenomena for the first time?

PICOT is a useful tool to assist in framing your research question using a 5-step process:

PICOT can be applied to a Randomised Controlled Trial (RCT), and has been found to result in better reporting quality when it’s used to structure research questions.[3]
eg Can a sleep intervention (I) compared with usual care (C), delivered to children aged 6-12 yrs with Attention Deficit Hyperactivity Disorder (P), improve the outcomes of child ADHD symptoms, behaviour and learning (O) at 3 and 6 months post randomization (T)?
PICOT can also be applied to a non Randomised Controlled Trial: In this case there is no comparator group required so the ‘C’ is dropped from the PICOT equation. The Intervention (I) refers to the dimension of health care under question.
eg Can procalcitonin (‘I’) taken within 48hours of onset of fever (T) detect bacterial illness (O) in children aged <3 years with a temperature >39 degrees Celsius presenting to paediatricians’ rooms (P) ?"

A useful article on PICOT (particularly pages.100-101) is:
Johnston, L and Fineout-Overholt, E. 2005, Teaching EBP: “Getting from Zero to One.” Moving from Recognizing and Admitting Uncertainties to Asking Searchable, Answerable Questions, Worldviews on Evidence-Based Nursing, 2(2), 98-102.

(iii) Further considerations
You may find that your area of research interest requires a stepped approach. For example if you design a RCT that examines the effectiveness of an education program to improve clinical care of eczema you may first need to identify physicians current practice and competency via a survey, secondly conduct a pilot to assess the feasibility and acceptability of the education session and then finally conduct the RCT.

[2] Hulley, SB. Designing Clinical Research, pp 20. © 2007 Lippincott Williams & Wilkins, Philadelphia, USA
[3] Rios LP, Ye C, Thabane L. Association between framing of the research question using the PICOT format and reporting quality of randomized controlled trials. BMC Med Res Methodol 2010 Feb 5; 10:11

Bookmark and Share