How to construct and formulate research questions

In many regards, the research question can be considered the most important part of a research project. Because of its centrality, it is a common source of struggle and anxiety for students as they learn to become junior autonomous researchers.
 
In this blog post, I tell you what the role of the research question is, why it is 300% worth spending time working on it and provide some ready-to-use advice on how to improve your research question, starting with its formulation.  

​Different types of research questions

To start, let’s distinguish between different types of research questions. The list, of course, is not exhaustive. I unpack here the research questions that the students I teach have to navigate in the development of their research projects in UK higher education.
 
1. “Working research questions”
 
Throughout a research project, we ask ourselves dozens of questions while conducting our inquiry. I call these questions “working research questions”.

​

---

For example, if you are investigating a research problem related to the shirts represented in this diagram, you may ask yourself questions such as:
-       “Who wears this kind of shirt?”
-       “What is the history of these shirts?”
-       “How do people have defined the word 'shirt'?”

Asking yourself these questions helps you figure out the bricks you need to put together to construct your project and make sure that you leave no stone unturned.

2. “Main research questions” 
This is a question that comprises a concept and that is pitched and formulated in a way that helps you to produce rich analytical work, rather than only description.

​For example, taking the example in the diagram used in this example, such a question could be “How do these unusual yet common shirts represent different types of social groups and symbolise different social statuses in fantasy movies since the 1980s?”
 
To answer this analytical question you need to answer several descriptive sub-questions:
- What is the history of these shirts?
- Who wears them?
- What do these shirts represent for different audiences?​

An analytical question will help you produce more analytical results (for more information on what I mean by “analytical”, read the blog post “What is analysis? Some tips to become more analytical”)
 
3. “Sub-questions”
 
Some research designs require the execution of distinct steps, each providing results that will then be analysed jointly, to be able to reach overall conclusions. This can be the case with mixed-method research (part of the research design is quantitative, part is qualitative) or multi-method (different methods are used and all can be either qualitative or quantitative). In this type of research design, different sub-questions may need to be investigated separately, as each constitutes a different piece of the puzzle that needs to be solved to answer the main research question.

​Students often mix working questions with sub-questions and think that every question they have asked themselves throughout their project is a sub-question and needs to be shared in their research. When this occurs in assignments, we often see the main research question followed by a list of five questions that are not central to the demonstration, but that the students have asked themselves at some point. Rather than strengthening the demonstration, it tends to dilute the impact of the main research question and distracts the student (and the readers). Indeed, this tendency reinforces the challenge that caused this confusion in the first place: the difficulty of identifying the main research question and establishing a hierarchy between different types of questions that we ask ourselves during research.

---

Not all the questions we ask ourselves throughout a research project need to be shared with the readers. Don’t forget that a piece of research is an argumentative discourse produced for someone else rather than a chronological archive of every idea you went through. In the same way that a music album is a finite product aimed for an audience to experience, a research paper is an intellectual product, a discourse aimed for an audience to learn from. While sub-questions are essential foundations to understanding the architecture of your project and need to be shared, you do not need to share all the working questions with the readers.

What is the role of the (main) research question?

In the rest of the blog post, I will focus on main research questions as this is what most students are asking me about. Regarding my students at the London School of Economics more specifically, these are the type of questions you are usually expected to formulate for 3000 word and 4000 word individual project assignments.
 
The main research question plays several important roles. This is why it is important to spend time working on it …. and a badly formulated research question often involves consequences rippling throughout the rest of the research design (aka shooting yourself in the foot!). A research question is a tool of research design:

• It guides you in the right direction and enables you to achieve the right level of ambition, that is to say, making the most out of your research project in a feasible way. I often use the metaphor of the archer shooting an arrow. If you shoot your research question arrow too high you might not be able to finish your project on time, if you shoot it too low, it does not enable you to achieve your project’s full potential.
• It is an accountability/self-coaching device that needs to be reassessed and actualised. Your first formulation of the research question is not a contract that you cannot depart from. That being said, you need to make sure that this departure (e.g. the reformulation or modification of your research question) is done in a conscious and informed way. Otherwise, you run the risks of just drifting away, which will end up with you losing time and straying off topic.
• It is a practical way of helping you bind (=narrow down and contain) your project and bring focus and coherence to your investigation.

 
The research question is also useful for the readers: make sure it synthesises your project’s lines of enquiry in one sentence that can strike their interest and act as a memorable takeaway!

​Formulating research questions

There are many ways to formulate research questions. Here are some tips to help you understand what is at stake in the formulation of the research question, and give you a steppingstone if you don’t know where to start.
 

'Students often approach the idea of “formulating” a research question as a way to polish it at the end of their project to make it “sound nice”. But there is more to it than that. The words we pick for our research question are more important than any other words in our research project: they are the magic ingredients that can take our research project to the next level. Taking the formulation of your research question seriously enables you to create a more coherent and impactful research design; it also lays a solid foundation for the rest of your project.
Let’s unpack how research questions are constructed through two examples (one rather qualitative and one rather quantitative). I have kept them abstract on purpose so it is easier for you to fill in the blanks with something related to projects that you may have in mind, rather than being distracted by the specifics of what these research questions may be about:

​How do [social group x] experience [something]?
Does [something] impact [social group]’s access to [something]?
 
In the following, I’ll unpack one by one the different types of words that constitute common research questions such as these two:

• How do you start your research question?​

How     does [social group x] represent [something]?
Does    [something] impact [a social group]’s access to [something]?
 
Notice the difference between how these two research questions start. The first one is open-ended (not answerable via 'yes' or 'no') while the second one is closed-ended (answerable via 'yes' or 'no'). This difference has big research design implications, both epistemological and methodological. For example, open-ended questions work well when you have a small sample with rich material that you can unpack analytically to show nuances, tensions and the like. In contrast, 'yes'/'no' questions will often end up leading to very flat results with a small sample. On the other hand, hypothetico-deductive designs often work well with closed-ended questions. Closed-ended questions can also help problematise assumptions that the literature takes for granted (if everyone agrees that something is a 'yes', and your exploratory work seems to suggest it might be a 'no', a 'yes' or 'no' question can potentially lead to very innovative/provocative results). So it is always a question of making sure the word that guides the nature of your question aligns with the other dimensions of your research design.

• The grammatical subject of your sentence

How do [social group x] represent [something]?
Does      [something] impact [social group]’s access to [something]?
 
The grammatical subject is the core of your research design. Make sure there is an alignment between what you chose as the grammatical subject and the rest of your research design choice: such as the empirical material and analytical framework you choose, as well as the literature you review.
 
For example:
 
How does [social group x] represent [something]?
 
Is different to:
 
How is [something] represented by [social group X]?
 
Here I am not talking about active style vs passive; the difference has implications in terms of what the project is about:
 
“How does the World Health Organisation represent childbirth?” reflects a project about international organisations, discourses of authority and policy-making etc while “How is childbirth represented by the World Health Organisation?” focuses on childbirth as a discursive and symbolic site, the history and geography of its representation of which international organisations represent the current instance your project aims to focus on.

• The verb characterising the phenomenon or process you will investigate

How do [social group x]  represent [something]?
Does      [something]        impact [social group]’s access to [something]?
 
This verb is the entry point to operationalisation, so you need to ask yourself: do you understand the process or phenomenon this verb entails and is your research design enabling you to empirically assess this process or phenomena?

• The question mark

How do [social group x] represent [something]?
Does      [something] impact [social group]’s access to [something]?
 
I encourage junior researchers to ask a question with a question mark. This will make it easier for you to identify whether your research question is actually a research question or something else (e.g. a hypothesis or variable in disguise). Make sure that your research question is different from the (formal or informal) hypothesis you may have and that there is not an answer already included in your question (that it is not a real question!).

Constructing research questions beyond formulation

​Here are a few criteria that can help you further brainstorm/self-assess your research question. Namely, your research question needs to be:

1. Well-bounded, not too narrow (“how does the coffee shop at my university advertise their mocha latte?”) nor too broad (“how do universities all around the world advertise merchandise?”). Include some elements of your case in the research question to help bound the claims you make.
2. Follow naturally from your introduction/literature review. If the rest of the introduction is well written, readers should be capable of guessing your research question without seeing it.
3. Formulated via concepts: concepts a) facilitate empirical operationalisation b) bring analytical depth to your project: e.g. it would be much harder for you to operationalise empirically and to bring analytical depth to your project if your research question is “How do British people portray Polish people in the UK?” than if your research question is “How do urban British youth represent eastern European immigrants?”.
4. Lead to empirical results via the methods you plan to use: make sure that your research question is answerable via the methods that you aim to use.
5. Able to be answered in one impactful sentence. That sentence will be the main takeaway of your project and you should make sure this sentence is explicitly written in your project and not let for your readers to guess.

 
To conclude, there is no cookie-cutter recipe. Each methodological tradition has its own specificities when it comes to what constitutes good research questions, which goes way beyond what a blog post (or my expertise!) allows. But I assure you that implementing the tips I give you in this blog post will help you on your research design journey 😉.