Vous êtes sur la page 1sur 5

How to...

How to…write a good

research question
Karen Mattick1, Jenny Johnston2 and Anne de la Croix3,4
Centre for Research in Professional Learning, University of Exeter, Exeter, UK
School of Medicine, Dentistry and Biomedical Sciences, Queen’s University Belfast,
Belfast, Northern Ireland, UK
LEARN! Academy, Vrije Universiteit Amsterdam, Amsterdam, the Netherlands
VUmc School of Medical Sciences, Amsterdam, the Netherlands

A research SUMMARY project sets out to answer. Most questions are typically more fo-
question is a This paper, on writing research research questions will lead to a cused on the immediate context:
questions, is the first in a series project that aims to generate new for example, the effectiveness of
question that a insights, but the target audi- an educational intervention in a
that aims to support novice
research project researchers within clinical educa- ence and the methodology will particular setting. Whether your
sets out to tion, particularly those undertak- vary widely. The term ‘evaluation ambition is for research or evalu-
answer ing their first qualitative study. question’ is used less commonly, ation, we hope that you will find
Put simply, a research question but the same principles apply. The this paper helpful for designing
is a question that a research key difference is that evaluation your own educational projects.

104 © 2018 John Wiley & Sons Ltd and The Association for the Study of Medical Education. THE CLINICAL TEACHER 2018; 15: 104–108

tct_12776.indd 104 2/28/2018 7:16:10 PM

INTRODUCTION researcher on a quest to identify research, we are often contribut- Ask yourself
or collect data that can be ana- ing a tiny step to the existing
what will

ost clinical teachers will lysed and interpreted, such that knowledge, rather than making
undertake a research it provides new insights. huge leaps. Small contributions happen if this
project, evaluate their are better than no changes at all. research is not
education practice or prepare a So what are the features of a So a long, specific question is done - does it
conference poster at some point good question? First and fore- likely to be preferable to a short really matter?
in their careers. Most clinical most, the question should focus vague question.
Who will
teachers will also explore the on an important topic. Ask
published literature on a topic yourself what will happen if this To illustrate these points benefit from it?
relevant to their practice: for research is not done – does it Table 1 provides some specific
example, a teaching innovation, really matter? Who will benefit examples, and in the next section
assessment strategy or student from it? Good questions are we discuss some key
support approach. These activi- often co-created with those who considerations.
ties can be enhanced by writing a may use or benefit from the
research question in order to findings. Look at the priorities CONSIDERATIONS:
think clearly about what you are identified by journals, funding RELEVANCE, ORIGINALITY
doing and why; however, not all bodies or priority-setting AND RIGOUR
clinical teachers have had exercises in the clinical educa-
training in how to write a good tion field to see what others Here, we highlight some things
research question, and this is not think are important questions.1 to think about as you develop
as straightforward as it may Sometimes new data might your research question, building
seem. Writing questions for prompt a research question: from on the points raised above. These
qualitative research, in particular, the findings of a national survey, analyses are not intended to be
might be new for many health for example. Other research undertaken in any specific order.
professionals. This article aims to questions are driven by a theory The considerations fall into three
share some tips and frameworks or hypothesis about what is main categories (Figure 1), which
that the authors (medical happening in practice. It is not can be thought of as broadly con-
education researchers who have uncommon for research questions cerning the relevance, originality
published quite extensively) have to start out by being quite ‘local and rigour of the research ques-
found useful, which we hope will and particular’, focused on the tion, and which are interrelated
support clinical teachers who are immediate educational context, and partly inspired by the UK
novice researchers in the educa- but they can often be developed Research Excellence Framework.3
tion sphere, and perhaps provide into questions with broader You may start with any one of
a helpful refresher for more relevance. the categories and revisit each
experienced researchers. one multiple times. In doing so,
Good research questions are you may change your research
WHAT IS A GOOD usually quite narrow or specific, question slightly or completely.
RESEARCH QUESTION? but often do not start out that Don’t be disheartened if you go
way. You might start with a through many iterations; time
A good question can make general theme or idea for spent on this process is always
people pause and see things in research (e.g. motivation for well invested.
a different way, or can motivate learning), which develops into a
them to learn more through more specific question over time Relevance
discussion, an internet search or (e.g. how do medical school Begin by identifying and ar-
literature review. When a child graduates engage with e-learning ticulating the important societal
asks their parents why the sea resources outside their working or practice problem that you
is blue, the parents may need to hours?). Novices often ask very wish to research. To consider
consult an information resource broad questions, but these are the significance of the research
before providing a well-informed unlikely to be answered in a question for different audiences,
answer! Thinking critically about short time frame and can lack try articulating the problem in
everyday taken-for-granted direction and impact.2 After multiple ways, perhaps working
assumptions or practice prob- developing a research question, with different stakeholders from
lems, to enable new potentially you will need to consider whether the clinical education field to
transformative viewpoints to it can be answered through the do this (Box 1). Talk about your
be articulated (a process called existing published literature or project idea in lay terms and
‘problematising’), is one of the whether new data must be observe people’s reactions with
most useful ways of generat- collected. Reviewing the litera- an open mind. The process will
ing research questions. A good ture is only manageable if the require you to think clearly and
research question will send the question has clear boundaries. In from different perspectives.

© 2018 John Wiley & Sons Ltd and The Association for the Study of Medical Education. THE CLINICAL TEACHER 2018; 15: 104–108 105

tct_12776.indd 105 2/28/2018 7:16:11 PM

Ensure that the Table 1. Examples of research questions in clinical education
research ques-
Research question Source Our comments
tion will lead to
original work How do people learn? Created by the authors Too broad! The question spans many disci-
that generates plines (education, psychology, sociology,
anthropology) and the answer is likely to
new insights
depend upon the ‘people’ studied.
Do postgraduate trainees Created by the authors Not aligned! The question demands a yes/
like lectures? An interview no answer, but the methodology will
study. provide words as data. The question also
focuses on learner satisfaction, but could
be developed to consider knowledge gain
or behaviour change.
How can we increase the Created by the authors This question is only important if there
number of applications for are insufficient high-quality applicants
a fixed number of medical to fill the places. Its importance could
school places? be increased significantly by focusing on
particular demographic groups who are
underrepresented at medical school.
What interactional struc- Rizan et al. (2014)14 The importance of this research question for
tures are used in feedback clinical education is clear. This is a ques-
sequences during general tion that is probably better suited to data
practice bedside teaching in the form of words.
How does cognitive empathy Smith et al. (2017)15 This long, specific question explicitly builds
specific to the doctor– on previous work and is probably better
patient relationship, as suited to quantitative data collection.
assessed in much of the In PICO terms, the population repre-
previous work, change sents medical students, the intervention
over the course of medical is a medical school, the comparison is
school? between students at the beginning and
end of medical school and the outcome is
cognitive empathy.

PICO: patient, intervention, control/comparison, outcome.

Relevance: what is the Originality: what is Another key consideration is to
importance or relevance already known about ensure that the research question
of the topic for the topic? Which debate
does it add to?
will lead to original work that
generates new insights and does
not duplicate previous research,
which can be determined through
a literature review. There are many
different approaches to literature
Research reviewing,4 so you can tailor this
question to the purpose. If you can find an
existing research study or review
article on your topic of interest
that answers (or partially answers)
the question you are trying to de-
fine, then you may not need to do
Rigour: is the question aligned the study you had first envisaged.
with the methods? Do the This is good news, as your time
research tools generate can now be spent extending that
appropriate data to answer the
question? work and contributing to knowl-
edge, rather than unintentionally
Figure 1. Three interrelated elements of research question development
duplicating what is already known.

106 © 2018 John Wiley & Sons Ltd and The Association for the Study of Medical Education. THE CLINICAL TEACHER 2018; 15: 104–108

tct_12776.indd 106 2/28/2018 7:16:11 PM

satisfaction, but it is possible Questions may
Box 1. Examples of stakeholders in clinical education to design research that consid-
also be shaped
• Teachers and learners in health care disciplines ers knowledge gain, behaviour
change or patient outcomes. by researchers’
o undergraduate
Bloom’s taxonomy is useful for preferred
o postgraduate thinking about the descriptors methodological
o continuing professional development used in relation to educational tools
outcomes.10 Readers might
• Patients and the public also be familiar with the PICO
o of all ages and demographic groups (patient, intervention, control/
comparison, outcome) frame-
o in hospital and community settings
work,11 with SPIDER (sample,
o individually and through patient groups phenomenon of interest, design,
• Policy makers evaluation, research type) pro-
viding a qualitative alternative.12
• Organisations The SPIDER framework can help
o universities to spell out the key elements of
a question for qualitative work.
o hospitals
• Health care systems Once you have started your
research project, revisit the
research question regularly. Some
Once you have identified the prob- Questions may also be shaped projects evolve and drift during
lem that you wish to research and by researchers’ preferred meth- the study, and ultimately do not
know what research has already odological tools, which are answer the research question. It
been undertaken, you will be able inevitably interlinked with the is easy to lose sight of the
to articulate the gap in the lit- way that they see the world:7 for question in the midst of the
erature that you wish to address. example, whether they have a all-absorbing research process.
Lingard’s paper on ‘Problem, Gap, qualitative or quantitative Different elements of research
Hook’, which suggests articulating orientation. Talk to fellow need constant realignment with
a current problem, highlighting researchers. What kind of study each other and Agee reminds us
an important knowledge gap and design would they anticipate that this is critical to the shaping
convincing the audience that this from this question? What kinds of of research studies, particularly
gap is problematic, is a useful way theories and methodologies qualitative studies.13 Keep the
of thinking about this.5 might be useful in answering this research question foremost in
question? Even towards the end your mind throughout the
Rigour of your research project, when process.
Ensuring that the research drawing conclusions, refer back
question and methods of data to the research question to CONCLUSION
collection and data analysis are ensure the coherence of the
aligned is a key element of rigour. study. A good research question takes
Research questions may suggest a time to create, but time in-
particular type of answer. For ex- TIPS FOR WRITING YOUR vested in this process is always
ample, some questions demand a RESEARCH QUESTION worthwhile.
yes or no answer, some require a
number or a ranking as an answer, When drafting your research REFERENCES
and some may be better answered question, there are some useful
with data in the form of words. frameworks that you can use 1. Dennis AA, Cleland JA, Johnston
P, Ker JS, Lough M, Rees CE.
It is worth noting that different to help you think clearly about Exploring stakeholders’ views of
people may be drawn to particular what you are doing and about medical education research priori-
types of question: some favour the components of your research ties: a national survey. Med Educ
‘What’, and some favour ‘Why’ or question. Readers might be 2014;48(11):1078–1091.
‘How’, and these preferences are familiar with Kirkpatrick’s model 2. Morrison J. Developing research
shaped by their knowledge and for evaluating learning and the questions in medical education:
prior experiences. Generally, ques- expanded version described the science and the art. Med Educ
tions focusing on experiences, by Barr et al., which consid-
viewpoints, group processes and 3. Anonymous. Research Excellence
ers different types of education
Framework. Available at http://
personal development tend to outcome that may be explored.8,9 www.ref.ac.uk/. Accessed on 8
lend themselves better to qualita- Much clinical education research January 2018.
tive research.6 to date has focused on learner

© 2018 John Wiley & Sons Ltd and The Association for the Study of Medical Education. THE CLINICAL TEACHER 2018; 15: 104–108 107

tct_12776.indd 107 2/28/2018 7:16:11 PM

Keep the 4. Wong G. Literature reviews: who 9. Barr H, Freeth D, Hammick M, 12. Cooke A, Smith D, Booth A.
research ques- is the audience? In: Cleland J, Koppel I, Reeves S. Evaluating Beyond PICO: The SPIDER
Durning SJ, eds. Researching Medical Interprofessional Education: A Tool for Qualitative Evidence
tion foremost Education. Chichester, UK: John United Kingdom Review for Health Synthesis. Qual Health Res
in your mind Wiley & Sons, Ltd; 2015: pp. 25–34. and Social Care. London, UK: 2012;22(10):1435–1443.
5. Lingard L. Joining a conversa- The British Educational Research
throughout the tion: the problem/gap/hook Association (BERA) and the United 13. Agee J. Developing qualita-
process tive research questions: a
heuristic. Perspect Med Educ Kingdom Centre for Advancement
reflective process. Int J Qual
2015;4(5):252–253. of Interprofessional Education
Stud Educ 2009;22(4):
6. Tai J, Ajjawi R. Undertaking and (CAIPE); 1999.
reporting qualitative research. Clin 10. Bloom BS, Engelhart MD, Furst EJ,
Teach 2016;13(3):175–182. Hill WH, Krathwohl DR. Taxonomy 14. Rizan C, Elsey C, Lemon T, Grant
7. Bunniss S, Kelly DR. Research of educational objectives: The classifi- A, Monrouxe LV. Feedback in
paradigms in medical cation of educational goals. Handbook action within bedside teaching
education research. Med Educ I: Cognitive Domain. New York, NY: encounters: a video etho-
2010;44(4):358–366. Longmans, Green and Co.; 1956. nographic study. Med Educ
8. Kirkpatrick Partners. The 11. Schardt C, Adams MB, Owens T,
Kirkpatrick Model. Available at Keitz S, Fontelo P. Utilization of 15. Smith KE, Norman GJ, Decety J.
https://www.kirkpatrickpart- the PICO framework to improve The complexity of empathy dur-
ners.com/Our-Philosophy/The- searching PubMed for clinical ing medical school training: evi-
Kirkpatrick-Model. Accessed on questions. BMC Med Inform Decis Mak dence for positive changes. Med
13 December 2017. 2007;7:16–21. Educ 2017;51(11):1146–1159.

Corresponding author’s contact details: Prof. Karen Mattick, Centre for Research in Professional Learning, University of Exeter, St Luke’s
Campus, Heavitree Road, Exeter, EX1 2LU, UK. E-mail: k.l.mattick@exeter.ac.uk

Funding: None.

Conflict of interest: All three authors teach on this topic to postgraduate students and supervise Masters and Doctoral-level project
students, giving feedback on these kinds of issues. KM is Senior Associate Editor for The Clinical Teacher.

Acknowledgements: We thank Esther Helmich, Aileen Barrett, Deirdre Bennett, Anu Kajamaa and Terese Stenfors for providing feed-
back on the draft manuscript. They, together with the authors, form a research network named the European Centre of Excellence in
Qualitative Study and Inquiry in Training and Education (EXQUISITE), led by Esther Helmich at the University of Groningen.

Ethical approval: No ethical approval was sought, as the work is based on secondary data and on our own experiences. No primary data
are presented.

doi: 10.1111/tct.12776

108 © 2018 John Wiley & Sons Ltd and The Association for the Study of Medical Education. THE CLINICAL TEACHER 2018; 15: 104–108

tct_12776.indd 108 2/28/2018 7:16:11 PM