Sep 25, 2026 06:08pm

By Benita Olivier

Under

23 min read

Finding your research topic with Consensus

You will be working on your research topic for between one and six years, and it will have an effect on the skills you acquire, the people you come across, the papers you produce, and sometimes on your future career as well; therefore, it is important to take the time to choose it carefully rather than deciding on it at the last minute.

The reassuring thing is that very few people begin with a research question. Instead, most postgraduate students start with an area of interest, such as a curiosity, a frustration from practical experience, or a feeling that something could be improved. During the early months, their task is to narrow down that general interest to a specific question which is both interesting, new, and feasible. This process involves clear, identifiable stages, and one can learn how to navigate them effectively.

A topic that is suitable will be one in which three conditions are met: it captures and holds your interest, it addresses a genuine gap in current knowledge, and you can actually finish it given the time, support and resources at your disposal. If any of these points are missed, you will eventually realise it, typically by the eighteenth month.

Tools powered by AI, such as Consensus, can greatly speed up the processes of searching, mapping and checking. However, they cannot carry out the thinking for you and indeed should not do so. The only way to get to an expert level in a subject is by reading extensively about it. This guide explains how you can use Consensus to work smarter at every stage, while still maintaining full control over your own judgement.

Just a quick note: I don't get commission from Consensus. I just really like the fact that they can make finding the gap more effective and fun! To access all the functions I show here, you'll need the Pro version, which will show you many more papers. The free version is useful to get an idea of the field, so do play around. Also, you can still use the various steps without the functions in Consensus - like how we've done it in the "good old days".

If you are embarking on a doctoral degree and you are considering using AI as part of your thesis or dissertation, check out our new book, "Still Yours: A Doctoral Researcher's Guide to AI". More info here.

The journey at a glance

This guide walks you through seven steps, from a broad interest to a question you can take to your supervisor.

Topic-finding is rarely a straight path, and it is completely normal to loop back to reading and mapping once someone asks you a hard question in Step 7.

A running example

To keep things clear, we will use one example throughout: how our beliefs affect our success, productivity, and behaviour. This topic is interesting and relevant to many people, and it seems well-studied at first. That makes it a good test case. If we can find a real research gap here, you can find one in your own field too.

Step 1: Start with what draws you in

You do not need any tools yet. All you need is a notebook and some honest reflection. Before you start searching online, take some time to think about three questions. They may seem simple, but your answers will help you stay focused when research becomes challenging.

1. What are you interested in? Curiosity is fuel. A topic that excites you will carry you through the long stretches of data cleaning and the third round of reviewer comments. Notice what you read about for fun, what makes you argue at dinner, and which problems from your work you cannot stop thinking about.

2. What are you good at, or working towards becoming good at? We tend to be drawn to what we do well. But a topic that stretches you is also a perfectly good choice, because a postgraduate degree is designed for learning. Just be honest about how big the stretch is.

3. Where do you want this degree to take you? Picture yourself two years after graduation. If you see yourself leading a team, working in a particular sector or building an academic career in a specific field, your topic can move you closer to it. Your reasons for studying will shape where you focus.

Write your interest statement

Pull your answers together into a single short paragraph. It does not need to be elegant. Here is ours:

I am fascinated by why some people seem to thrive under pressure while others with equal ability struggle. I suspect the beliefs we hold about ourselves play a big role in how productive and successful we are. I work with postgraduate researchers and would love my research to help them work more sustainably.

Notice what this paragraph includes and leaves out. It mentions a broad topic (beliefs and success), a group the writer cares about (postgraduate researchers), and a value (sustainability). It does not have a research question yet, and that is fine at this stage.

Remember: start with a general area, not a specific question. If you rush to a question before understanding the field, you might end up working on something that was already answered years ago.

Step 2: Get oriented and learn the field's language

Read about the field before you choose your topic. You need to know what exists before you can see what is missing.

When you first start reading about a new field, focus on getting oriented. Try to understand the main debates, key studies, recent reviews, and especially the terms that researchers use.

Learn the vocabulary

Researchers rarely publish in everyday language. They use field-specific terminology. Our broad interest splits into several established constructs, each with its own theory, measures and body of evidence:

Self-efficacy: belief in your capability to succeed at a specific task (Bandura's social cognitive theory)
Growth mindset: the belief that abilities can be developed through effort and strategy (Dweck)
Locus of control: whether you see outcomes as driven by your own actions or by outside forces
Impostor phenomenon: the persistent belief that your success is undeserved and you will be "found out"

If you search for "beliefs and productivity," you will only find a small part of the research. Using the right terms will give you a much better overview. Keep a list of the terms, synonyms, and outcome measures you come across. This glossary will be very useful for your literature search and your review later on.

Let the field surprise you

This stage is also where you test your assumptions. For example, you might think self-efficacy strongly affects work performance, and many studies do show a positive link. However, a well-known meta-analysis found that self-efficacy's effect is much smaller when you consider personality, mental ability, and experience (Judge et al., 2007). This does not mean you should drop the topic. These details can help you ask a better research question.

Consensus step-by-step: getting oriented

1. Start broad and neutral. Ask an open question such as "What is known about the relationship between self-beliefs and work performance?" Avoid leading questions like "Does self-belief improve productivity?", which invite confirmatory evidence.

2. Choose your search mode. Quick gives a fast overview. Pro and Deep analyse more papers and produce richer summaries, which is ideal when you are mapping an unfamiliar field.

3. Hunt for reviews first. Ask "What systematic reviews and meta-analyses exist on growth mindset and performance in adults?" Reviews are your fastest route to understanding a field's shape.

4. Ask for a research timeline. Consensus can build a timeline from your search. Seeing how a field developed (for example, from early mindset studies to large field experiments and replication debates) helps you understand why today's questions look the way they do.

5. Save as you go. Add key papers to a collection in your Consensus library. You will come back to this collection in Step 3.

Stay in control: Consensus can point you to the right papers in minutes. Reading them is still your job. The summaries help you decide what to read; they are not a substitute for reading it. This is the stage where you build the expertise that will carry your whole degree.

Step 3: Map the evidence with the Research Gaps Matrix

See where research has clustered, and where it looks thin.

Once you understand the key terms, you can start mapping the field. In the past, this meant spending weeks reading reviews and making spreadsheets to track who studied what, with which groups, and how. The Research Gaps Matrix in Consensus makes this process much faster.

What the matrix does

Ask Consensus about the research gaps in an area, and it builds a coverage grid from the papers retrieved for your question. Sub-topics run down one side, research dimensions (such as population or study design) run across the top, and each cell shows how many papers sit at that intersection. Dense cells show where attention has gone. Sparse or empty cells point to questions that might be worth investigating.

The matrix is also transparent. You can open any cell to see which papers are in it and read the exact passage that put each paper there, so you are never asked to trust a number blindly.

What it might look like for our example

Imagine you ask about gaps in research on self-beliefs and productivity, comparing belief types with populations. A simplified version of the pattern you might see is below.

The story jumps out. Growth mindset research is heavily concentrated in school and undergraduate settings, while doctoral researchers are much less studied. That matches what the literature itself says: a 2026 study describes the application of growth mindsets in doctoral education as having "received limited attention" (Wang et al., 2026). We have our first candidate gap.

Remember, if you don't have access to Consensus Pro, you can create a similar matrix offline.

Consensus step-by-step: using the Research Gaps Matrix

1. Ask for the gaps. Try a prompt such as "What are the research gaps in the relationship between growth mindset, self-efficacy and productivity?" or be explicit: "Show me a research gaps matrix comparing belief type to population."

2. Give it a good stare. Before zooming in, notice where the dense clusters and the empty spaces are. What surprises you?

3. Hover over a cell. You will see the category, its coverage level and the papers inside it, listed by author and year.

4. Open a paper. Click a paper and it opens beside the matrix, with the passage that justified its placement highlighted.

5. Check the Evidence tab. This collects every quote the matrix used, grouped by cell and labelled by whether it came from the abstract or the full text. Use it to check the reasoning rather than treating a count as a conclusion.

6. Map your own library. Run the same analysis on the collection you saved in Step 2, for example: "Show me a research gaps matrix comparing population to study method in this collection so I can see what I am missing." This reveals blind spots in your own reading, not just in the field.

Try swapping the dimensions. Comparing belief type with study design rather than population can reveal a different kind of gap. In our example, you might notice that most studies are cross-sectional surveys, with far fewer long-term or intervention studies. Several authors say the same: rigorous longitudinal field studies of self-efficacy and job performance are described as rare (Carter et al., 2018).

Step 4: Verify the gap

An empty or sparse cell is just a possible gap, not a confirmed topic.

This is the step that separates a good topic from an embarrassing viva moment. An empty cell only means that the papers you found did not cover that area. It does not prove that no research exists. The lack of coverage could be real, or it could be due to how you searched, the terms you used, or research published under a different name.

Know what kind of gap you are looking at

Not all gaps are the same. Miles (2017) describes seven types of research gap. Here is how each might appear in our example:

Type of gap	What it means	In our example
Evidence gap	Findings contradict each other	Two major 2022 meta-analyses reached different conclusions about whether growth mindset interventions improve achievement (Macnamara & Burgoyne, 2022; Burnette et al., 2022)
Knowledge gap	The desired knowledge does not yet exist	How beliefs relate to both productivity and well-being over a doctoral journey
Practical-knowledge gap	Practice differs from what research supports	Mindset training is widely promoted despite contested evidence
Methodological gap	A different method could shed new light	Most studies are cross-sectional surveys; few follow people over time
Empirical gap	Findings or propositions need testing or validating	Workplace findings on self-efficacy have not been tested in research-degree settings
Theoretical gap	Theory is missing or underdeveloped	How beliefs translate into day-to-day research behaviours is poorly specified
Population gap	A group is under-represented	Doctoral researchers, particularly those studying part-time alongside work

Check before you commit

Work through these checks for every candidate gap:

Investigate further with the Deep function in Consensus. In the matrix, an empty cell is marked "Potential gap". Choose Investigate further with Deep and Consensus turns that exact row-by-column intersection into a focused search. See what comes back.

Search beyond one tool. Repeat the search in the main databases for your discipline (for example, PsycINFO, Scopus or Web of Science) using the vocabulary from Step 2. Your subject librarian is a superpower here.
Look for work in progress. Check registries such as PROSPERO (for systematic reviews), ClinicalTrials.gov (for trials) and OSF Registries (for pre-registered studies). Someone may already be filling your gap.
Search theses and dissertations. Much doctoral research sits in university repositories and dissertation databases.
Mine the "future research" sections. Recent papers in your area often tell you directly what the authors think is missing. Recurring suggestions are a strong signal.

What verification looks like in practice

In our example, investigating the "growth mindset × doctoral researchers" cell turns up a recent mixed-methods study of 379 doctoral students that found stronger growth mindsets were associated with higher research skills, self-efficacy, productivity, and satisfaction (Wang et al., 2026).

This is a good outcome. That study measured students at one point in time and explored their experiences through interviews. It tells us beliefs and productivity travel together; it cannot tell us how they change across a degree, or what happens to well-being along the way. Another study even found that doctoral students' research self-efficacy and publication numbers were negatively related to mental well-being (Zhang et al., 2022). Our gap is becoming clearer.

Remember: just because you do not find something in one search does not mean it does not exist in the literature. Treat every gap as something to test, and keep track of where you searched. This record will help you show evidence for your gap in your proposal.

Step 5: Ask whether the gap is worth filling (and whether you are the one to fill it)

Not every gap is worth turning into a thesis. Some gaps exist for valid reasons.

You now have one or more verified candidate gaps. Before falling in love with any of them, put each through two tests: is it worth filling, and can you fill it well?

Is it worth filling?

Will it create impact? Ask the "so what?" question out loud. Would the answer help solve a real problem, bust a myth, change a practice or inform the next wave of research? Doctoral attrition and researcher well-being are pressing concerns, so a study showing how beliefs shape sustainable productivity in doctoral researchers has a clear audience: students, supervisors and graduate schools. Even when a topic has been explored before, we usually still don't know much.

Is there a reason no one has studied this gap? Sometimes gaps exist because the research would be unethical or impossible to measure. It is a good idea to ask someone experienced before you commit to the topic.

Are you the one to fill it?

Is there a supervisor with expertise in your niche? Look for people who publish in the area (Consensus will show you who keeps appearing in your searches). If the perfect expert is not at your university, a supervisor with strong methodological expertise plus an external co-supervisor or adviser with topic expertise can work beautifully.

Could you join an existing research group or project? Slotting into an established project often gives you access to participants, data and colleagues, and makes the journey much less lonely. It is not essential, though. Plenty of excellent theses are completed without one.

Are the resources there? Think beyond tuition fees to the cost of the research itself: equipment, software, participant incentives, transcription and travel. A survey study costs far less than one requiring specialist equipment or laboratory tests.

Is it doable in your time frame? A longitudinal study following doctoral researchers for three years might be the perfect design, but it will not fit a one-year master's. Be realistic about access to participants, data and ethics approval.

Does it fit where you are heading? Return to your interest statement from Step 1. Does this gap still move you towards the future you described?

Compare your candidates side by side

If you have two or three candidate gaps, a simple comparison helps. Rate each from 1 (low) to 3 (high):

Criterion	Gap A: growth mindset and sustainable productivity in part-time doctoral researchers	Gap B: impostor beliefs in health professionals
Interest to me	3	2
Verified gap	3	2
Impact and need	3	3
Supervisor expertise available	2	1
Feasibility (time, access, cost)	2	2
Fit with career goals	3	2
Total	16	12

The scores are a conversation starter, with yourself and with your supervisor.

The principle to remember: the best topic is not the most impressive one. It is the one that is worth doing and that you can finish well, while still having a life. Control what you can control, and choose accordingly.

Step 6: Turn your gap into a focused research question

A gap describes what is missing. A research question says exactly what you will find out.

With a verified, worthwhile gap in hand, you can finally write your question. Expect several drafts. This is how ours evolved:

Broad interest: How do our beliefs shape our success?
Narrowed area: How do growth mindset beliefs relate to productivity in doctoral researchers?
Focused question: Among part-time doctoral researchers in employment (P), how are growth mindset beliefs at the start of their studies (E) associated with research progress and well-being over the first two years (O)?

Each draft becomes more specific about who is involved, what is being studied, and which outcomes matter. The final version also includes a time frame that matches the gap we identified.

Use a framework to structure it

Frameworks ensure your question includes all essential parts. Our example uses PEO (Population, Exposure, Outcome) because growth mindset is a naturally held belief rather than an intervention we deliver. If we wanted to test a mindset workshop instead, PICO (Population, Intervention, Comparison, Outcome) would fit better. If we wanted to explore how doctoral researchers experience self-doubt, SPIDER would suit a qualitative design. The companion guide From Framework to Prompt explains each of these frameworks in detail.

Test it against FINER

The FINER criteria (Hulley et al., 2013) offer a quick health check for any research question:

Feasible: Can you recruit enough participants and complete it within your time and budget?
Interesting: Does it interest you, your supervisor and your field?
Novel: Does it confirm, refute or extend previous findings? (Step 4 is your evidence here.)
Ethical: Can it be done without undue risk or burden to participants?
Relevant: Will the answer influence practice, policy or future research?

Consensus step-by-step: stress-testing your question

1. Run your focused question as a search. If Consensus returns a stack of studies that answer your question exactly, you have learnt something important, and it is far better to learn it now.

2. Check each component. Search the parts separately (for example, "How are growth mindsets measured in adults?") to find validated measures and see how other researchers have handled similar designs.

3. Rerun the matrix on your narrowed question. A research gaps matrix on the narrowed topic, comparing outcome with study design, shows whether your planned approach fills a thin cell or joins a crowded one.

4. Ask for a flowchart. A flowchart of how others designed similar studies can help you start sketching your own methods.

Stay in control: feel free to ask AI to suggest alternative phrasings of your question. But the final wording should be yours, and you should be able to defend every word of it.

Step 7: Test it with others

A topic that survives a few good conversations is a topic worth building on.

You have put a lot of thought into your topic. Now, let others ask questions about it. Their feedback will reveal assumptions you might not have noticed, and that is exactly what you want at this stage.

Write a one-page topic brief

Before meeting your supervisor, capture your thinking on a single page:

The area: your interest statement, refined
The gap: what is missing, what type of gap it is, and the evidence that it exists (including where you searched)
The draft question: structured with a framework
Why it is worth doing: the impact and who will benefit
Feasibility notes: participants, data, resources, time and ethics
Your open questions: what you are unsure about

Having a brief like this changes a vague conversation into a productive meeting.

Who to talk to

Your supervisor (or prospective supervisor) is your most important sounding board. They will know whether your gap is real, whether the design is realistic and whether the topic fits the requirements of your programme.

Peers and research groups will ask valuable questions, and talking about your topic clarifies things in your own head. Present your brief at a lab meeting, a journal club or a writing group.

The people your research is about can tell you whether your question addresses something that is important to them. In health and social care research, patient and public involvement is expected practice. In our example, a short conversation with a few part-time doctoral researchers might reveal that "progress" means something quite different to them than it does to a graduate school.

Expect to revise. Many strong topics go back to Step 2 or Step 3 at least once. Revisiting the drawing board is how good research questions are made.

What if your topic has already been chosen for you? Many funded studentships and clinical research posts come with a defined project. The steps in this guide still apply. Use Steps 2 to 4 to understand the evidence around your project, then look for the angle, sub-question or population within it that can become distinctly yours. Owning a piece of a larger project is excellent preparation for an independent research career.

Your topic-finding checklist

Before you search

I have written an interest statement that captures my area, the people I care about and where I want this degree to take me.
I have built a glossary of the terms, constructs and measures researchers use in my field.
I have read key reviews and seminal papers myself, not just AI summaries.

While you map and verify

I have used the Research Gaps Matrix to see where the evidence clusters and where it looks thin.
I have opened the cells that interest me and checked the papers and quotes behind them.
I have treated every sparse cell as a candidate and verified it through Deep, other databases, registries and thesis repositories.
I can name the type of gap I am addressing and have kept a record of where I searched.

Before you commit

My gap passes the "so what?" test and I know who will benefit from the answer.
I have checked supervision, resources, time frame, access and ethics.
My question is structured with an appropriate framework and passes the FINER check.
I have tested my topic brief with my supervisor and at least one other person.
I have disclosed my use of Consensus and other AI tools in line with my institution's guidelines.

A final thought

The gap you find is only as good as the thinking you bring to it. Consensus can map a field in minutes, show you where the evidence runs thin and help you check every claim along the way. That is a remarkable head start. But deciding which gap is worth years of your life, and shaping a question you can defend, remains wonderfully and stubbornly human work.

Use the tools to go faster. Use your judgement to go in the right direction. And enjoy the process: this is the stage where your research becomes yours.

Start your search: try "What are the research gaps in [your topic]?" in Consensus and see what the matrix reveals.

If you want to do better research using AI but feel a bit uncertain about how to use it safely, check out our new book, "Still Yours: A Doctoral Researcher's Guide to AI". More info here.

References

Burnette, J. L., et al. (2022). A systematic review and meta-analysis of growth mindset interventions: For whom, how, and why might such interventions work? Psychological Bulletin. https://doi.org/10.1037/bul0000368
Carter, W., et al. (2018). The effects of employee engagement and self-efficacy on job performance: A longitudinal field study. The International Journal of Human Resource Management. https://doi.org/10.1080/09585192.2016.1244096
Hulley, S. B., Cummings, S. R., Browner, W. S., Grady, D. G., & Newman, T. B. (2013). Designing clinical research (4th ed.). Lippincott Williams & Wilkins.
Judge, T. A., et al. (2007). Self-efficacy and work-related performance: The integral role of individual differences. Journal of Applied Psychology. https://doi.org/10.1037/0021-9010.92.1.107
Macnamara, B. N., & Burgoyne, A. P. (2022). Do growth mindset interventions impact students' academic achievement? A systematic review and meta-analysis with recommendations for best practices. Psychological Bulletin. https://doi.org/10.1037/bul0000352
Miles, D. A. (2017). A taxonomy of research gaps: Identifying and defining the seven research gaps. Doctoral Student Workshop: Finding Research Gaps, Research Methods and Strategies, Dallas, Texas.
Wang, F., et al. (2026). Growth mindsets and research development: A mixed-methods study among doctoral students. Studies in Higher Education. https://doi.org/10.1080/03075079.2026.2675462
Zhang, F., et al. (2022). Social predictors of doctoral student mental health and well-being. PLoS ONE. https://doi.org/10.1371/journal.pone.0274273

***

Consensus feature descriptions are based on the Consensus announcement of the Research Gaps Matrix (Turn potential gaps into focused research questions, 22 September 2026).

I do not earn a commission from Consensus to refer to them in this post.

The opinions expressed in this blog post are not necessarily those of my employers.