How to compose the best research question for your study

If you’re having difficulty moving forward with your research, I want you to ask yourself a simple question: how clear is your research question?

No, really, is it crystal clear? 

Because so often when people come to me stuck with their analysis, the real problem is that their research questions just aren’t clear enough.

Why your research question is the most crucial part of your project

Of course, there are lots of elements that need to be right for a research project to succeed. Your study design, your data collection and analysis methods, not to mention the actual writing of the paper! 

But none of this is possible without a research question that sticks. That’s because the clearer your research question is, the easier it is to plan and do your analysis (which conveniently are the next steps on my one-to-one programme!). Failing to plan is planning to fail, as they say!

On the flip side, if you’re not sure what you want to find out then it’s pretty hard to figure out what analysis you should be doing. You’ll struggle at every step of the process without a clear research question. 

This is why I always look at the research question first when I start with new clients. Once you’re in the middle of your project (or even in the middle of the planning stage), it’s easy to lose sight of those details. We usually discover that, by clarifying the purpose of the study, questions about data collection and analysis can be answered much more easily. 

Let’s look at some examples

Can you spot which of these are good or bad examples of research questions?

  • What factors affect outcomes in Type 1 diabetes patients in hospitals?
  • Does gender affect cardiometabolic outcomes?
  • Is infant mortality lower when labour is artificially induced at 37 weeks in pregnant women who have had a previous still birth instead of waiting for spontaneous induction?
  • Which combination of analgesics is best for reducing migraine headaches in patients for whom a single analgesic is ineffective?

Broad statements aren’t good enough

The first two examples need more detail.

If you retrieved all of the data on outcomes in Type 1 diabetes patients in hospital then you’d drown under a giant spreadsheet!

The last two questions are much clearer.

How to write the best research question for your study

Even if you feel sure that your question is right, it’s important to revisit it regularly with fresh eyes. Often writing your research question isn’t just a one-off job. We have to go back and tweak and refine as we realise that there’s still uncertainty in there.

Step 1. Think about your ‘why’

What do you want to find out? Why? What will it change?

I usually recommend trying to address a specific clinical question in your research question. Then your research will help to move healthcare forward.

Step 2. Write down your first draft

I know it sounds basic, but you’d be surprised how many researchers I meet who haven’t written down their research questions! 

Writing out your research question will make it much easier to notice the gaps and uncertainties. It is also much easier to think objectively about something written down than something in your mind. 

Avoid jargon as much as you can. I have always found that simplicity is the best way to find clarity! Start by writing the question down as you would describe it to a friend and then tweak from there. 

Step 3. Get specific.

Once you’ve got your ‘why’ and your first draft, use the PICO framework to frame your research questions and make sure they’re detailed enough. PICO covers the following:

Population – what type of patient or population is your study looking at?

Intervention – what therapy or process are you testing?

Comparator – what’s the control group, or what are you testing your intervention against?

Outcome – which primary and secondary endpoints are you assessing?

Each of these points needs to be clearly addressed by your research question. You can also add detail relating to timing/duration and the study type, so you may see this referred to as PICOS instead where the S refers to study type.

Quick task: Go back to the last two examples above and see if you can identify the PICO in each of them.

Step 4. Get a fresh pair of eyes on it.

There’s nothing like a peer-review to raise questions that you haven’t thought about yet. Sometimes it can feel a little nerve-wracking to open ourselves up to potential criticism. Positive criticism is one of our best learning tools though as everybody approaches a problem for a different angle.

How many research questions can you have?

There are no rules around this.

You want to make sure that you answer your questions using robust methods and it can get harder to do this as the number of questions increases or as they start to diverge in their intent. It’s also important to consider your sample size, otherwise you’ll be under-powered to answer lots of different questions. So it’s basically a bit of a balancing act!

The most important thing is to identify your primary research question. This is the thing that is most important for you to find out, and the main reason why you’re doing your study.

In nearly all circumstances, it’s best to select only one primary research question. This is then the basis for choosing your primary outcome and primary analysis. All of your other research questions then become secondary or exploratory research questions.

Research means constantly asking questions

I truly believe that a career as a researcher is an opportunity for lifelong learning. The research questions you choose will shape your career and allow you to explore the topics that really matter to you and to society. That’s why it’s so important to get them right and to keep asking more questions every time you start a new study.

The good news is, that if you can make it a habit to consistently review and refine the questions you’re asking, you’ll be able to enjoy a hugely satisfying and successful career in health research!