Framing your research question

Dr. Sharif Abbasi

 Introduction
• Your research question may evolve, or even
change throughout your research but it is
essential to maintain focus on your research
question at all times. Because of this, it is
important to start thinking about your research
question at an early stage.
 What makes a good research question?

• 1. The 'so what test'

• A good research question should pass the "so what"
test. In other words your question needs to be worth
answering.
• 2. Narrow rather than broad
• The focus of your question should be narrow rather
than broad. One reason for this is to ensure that the
question is answerable within the time allotted.
 What makes a good research question?

• 3. Answerable
• This is an obvious point but you do need to ensure
that the question is answerable given your
background and the research methods available to
you.
• 4. Original
• Your question cannot have already been answered.
 What makes a good research question?

• 5. Symmetry of outcomes
• Your research question should produce significant
findings for all possible outcomes.
• 6. Interesting to you
• You may not have the luxury of choosing a question
that directly appeals to your interests. However if you
do, then choosing an interesting question will help
you sustain motivation over the course of your
research.
 Implications

• It was stated that a symmetry of outcomes is important for

your research question, that is your question should produce
significant findings for all possible outcomes. 
• It's tempting not to think about the implications of your
research question until after you have gathered all your data. If
you do this, there's a serious risk that you won't be able to find
an interpretation for your results, or that you'll realise you
haven't asked the right questions.
• It's much wiser to think about the implications of your
research question while you are planning your study.
 Implications

• Example
• Suppose that you want to see whether giving office staff a plant to
care for in their part of the office makes them happier. You can try to
draw up a table of all the possible findings as follows:

• Finding A: All workers are happier.

• Finding B: Some workers are happier, and none are less happy.
• Finding C: Some workers are happier, and some are less happy.
• Finding D: No workers are happier, and some are less happy.
• Finding E: All workers are less happy.
 Implications

• Asking questions
• You can now ask questions, such as what the implications
would be of each finding. If one of the findings has no
implications, what does this tell you about your research
design? What would you do if you completed the study and
ended up with the finding that had no implications?

• You can also see that some of the findings are poorly defined
because they don't distinguish between whether workers are
only slightly happier, or very much happier.
 Implications

• Alter the research design

• The next step would be to change the research design so that it produces
outcomes that are more clearly interpreted.
• Usually the answer to a research question gives you insights which raise a
set of new, follow-on research questions. For example, imagine that you
find that all workers are happier as a result of being given a plant to care for.
This raises questions, such as why they become happier, how much happier
they become and what would happen if you gave them something other than
a plant such as a picture.
• Each answer also has implications. Some answers have clear practical
implications. In this example, the answers have implications about how to
make the office staff happier. Other answers may have theoretical
implications or methodological implications. The best research answers
have implications for practice, theory and methodology.
 Implications

• Practical implications
• Practical implications can raise difficult ethical questions. For example, a
piece of research into the psychology of stress, conducted with the
intention of reducing stress in the workplace, may be used by unethical
managers who want to manipulate their staff through stress. For this
reason, many social science researchers are reluctant to focus on practical
implications of their findings and focus instead on the methodological and
research implications, which will help researchers gain a better
understanding of human behaviour.
Exercise
Try writing a research question of your own. Then write the possible answers
you might find and what the implications would be for each finding.
 Thinking about your research question

• Most research students begin their research with a lot of enthusiasm and
energy and most are aware that they have much to learn. Choosing a good
research question looks on the face of it relatively simple and an area
where they can do something immediately. The reality, however, is that
choosing a good research question is actually much more difficult than it
appears.
• Chris and Jamaal discuss their ideas for research questions
• Chris: I've got this really great research question. What are the main
problems affecting secretaries? That should find all sorts of interesting
things.
• Jamaal: I started off with a question a bit like that, but my supervisor
wanted me to go for something much smaller. I'm doing a comparison of
office based problems identified by observational and interview-based
methods. It sounds really boring. I just hope I can keep going until the end.
 Thinking about your research question: seven years later

• Chris: How did the interview go?

• Jamaal: Yeah, fine thanks. But the novelty of working with journalists
wears off pretty quickly. They all get really excited about the way that my
work identified an easy solution to a major problem that they all missed.
They're not so interested in the stuff that I found most interesting. How's it
going with you?
• Chris: Well. There was all that trouble with the office staff complaining
about me calling them 'secretaries'. The Dean managed to sort that out.
• Jamaal: Yeah.
• Chris: After I wrote a letter of apology. I've finished the data analysis now
and 90% of them complained about people dumping work on them with
unreasonably short deadlines.
• Jamaal: And?
• Chris: Well that's just it really. Yeah. I hope to get something better out of
my next assignment.
• Jamaal: Yeah.
 Thinking about your research question
• Chris used a research question whose phrasing was likely to
antagonise the participants. He also failed to think about what
he would do if his results were trivial. Both these problems
could have been avoided by piloting his study before
beginning the main data collection.
 Agendas and philosophy 1
• There's an old saying that to someone with a hammer, every problem looks
like it needs to be hit. Every research question involves preferences,
decisions and assumptions about what to research and how to research it.
Some of these preferences, decisions and assumptions come from the
researcher as an individual; others reflect the researcher's social context.
They can have far-reaching implications, which can easily be missed, as
discussed elsewhere in this course.
 Agendas and philosophy 1
• Exercise
• Here are an example showing interactions between a research question and
its social context.
 Agendas and philosophy 1
• Decisions and assumptions
• Often there are no right or wrong answers about which decisions and
assumptions to make. Here are some, further, classic examples of
interactions between research questions and their social context.
 Agendas and philosophy 1
• Decisions and assumptions: What I will research

• Many researchers focus on ‘hot’ topics - the ones where the most
productive, high-profile research is currently happening. These topics
usually have a lot of research funding available and plenty of venues for
publication. However, choosing a hot topic means that you’re following the
fashion of that research community. It can also cause problems for you
when the hot topic stops being fashionable. What do you study next and
how long will it take you to acquire the skills to study the next topic?
 Agendas and philosophy 1
• Decisions and assumptions: Method to use

• Choosing a research method is a trade-off between realism and precision,

heavily influenced by practical considerations, for example, whether you
happen to have social contacts in a group you want to study. There are
also choices about what you consider to be a trustworthy or important
source of evidence. Some research communities consider people’s spoken
words to be the best source of evidence. For instance, if you are
researching urban legends, then this is the most obvious place to start, and
it makes interview-style approaches the most suitable. Other research
communities focus on non-verbal data (e.g. ergonomics) or use verbal data
as a starting point for using other methods (e.g. classic psychology
experiments into verbal priming).
 Agendas and philosophy 1
• Decisions and assumptions: Funding effects
• It’s difficult to do research without funding and there are numerous funding
bodies. Some have a clear agenda of fixing a problem. Some problem-
fixing agendas are supported by almost everyone, for example, medical
research into cures for malaria. Others are not so straightforward. For
instance, funding research to tackle abandonment of old people by their
families makes the implicit assumption that old people should be cared for
by their families, as opposed to investigating different ways in which old
people can be cared for. It’s not always a simple issue of right and wrong:
‘family support’ is generally agreed to be a good thing, but so is ‘choosing
the right type of care for the individual’.
• Another issue is that the funder’s expectations can put you under subtle
pressure. For example, if you are being funded by a company, they will be
hoping your findings will not reflect badly on them.
 Agendas and philosophy 1
• Decisions and assumptions: Where I will publish
• If you want to build a research career, you’ll need to publish in the leading
journals for your area. The number of leading journals in any area is
small, and their editors and reviewers are part of a research community
with assumptions and values. If your articles go against those assumptions
and values, you may have difficulty getting published.