Should You Start With a Topic, a Problem, a Question, or a Gap?

What Should Come First When You Start Research?

One piece of advice tells you to choose a topic first. Another tells you to identify a research problem. Someone else insists that serious research must begin with a gap in the literature. Then you encounter a research guide that starts by asking you to formulate a question.

These instructions can sound contradictory because they describe different ways of entering the same developmental process. Research projects do need topics, problems or rationales, focused questions or aims, and some defensible relationship to what is already known. They do not necessarily appear in that order.

The more useful question is therefore not “Which one must always come first?” but “Given what I know right now, where can I start most productively, and what must I establish next?”

There Is No Universal First Step

Each Starting Point Gives You Something Different

Starting With a Topic: Useful When You Know the Area but Not the Study

A topic is often the easiest entry point when you know what interests you but do not yet know exactly what needs investigating. You might begin with coastal adaptation, adolescent sleep, artificial intelligence in assessment, workplace surveillance, or antimicrobial resistance.

Some university research guidance explicitly teaches this general-to-specific progression. Monash University's Psychology Research Portal, for example, begins its research process with identifying an area of interest and a specific topic, followed by investigating that topic and developing a research question. George Mason University's Writing Center similarly recommends choosing an interesting general topic, conducting preliminary research, and then developing questions.

The strength of starting with a topic is freedom to explore before committing prematurely. The weakness is that a topic alone provides little justification or direction. “Artificial intelligence in assessment” tells you where you are looking, not what needs to be discovered.

If this is where you are, the immediate task is not to invent an impressive title. Begin developing the broad interest into a focused research topic, while reading enough to understand the conversations already occurring around it.

Starting With a Problem: Useful When Something Clearly Needs Investigation

Sometimes research begins because you encounter a concrete difficulty, unexplained pattern, uncertainty, or recurring decision for which existing knowledge appears inadequate.

A clinician may notice that patients repeatedly abandon a service despite apparently needing it. A school may introduce an intervention but observe large differences in uptake. An engineer may encounter a recurring failure under conditions not adequately represented by existing models.

Starting with a problem gives the research immediate practical relevance, but it creates another danger: assuming that your initial interpretation of the problem is correct. What appears to be a motivation problem might be an accessibility problem. What looks like technology resistance might involve workflow, training, incentives, or something else entirely.

Recent Monash guidance on developing needs statements makes a similar distinction between noticing a broad problem area and establishing that a genuine unmet need exists. It recommends scoping the problem and testing assumptions rather than immediately treating an initial impression as a confirmed need.

So when you begin research with a real-world problem, use preliminary investigation to establish what the problem actually is before building the study around your first explanation of it.

Starting With a Question: Useful When Curiosity Is Already Specific

Sometimes the first thing you have is genuinely a question. Perhaps you notice that two apparently similar communities respond differently to the same intervention and immediately wonder why. Or a seminar leaves you asking whether a familiar theory still predicts behavior under a new condition.

A question is productive because it naturally directs attention toward what you need to find out. Research questions help focus inquiry, and a strong question should be clear and sufficiently focused for the investigation it is meant to guide.

But an early question should be treated as provisional. You may discover that researchers have already answered it, that its assumptions are unsupported, that important concepts are poorly defined, or that answering it would require evidence you cannot obtain.

Monash's guidance describes good research questions as formed and iterated rather than simply found. That captures an important feature of the process: beginning with a question does not mean preserving its first wording.

Starting With a Gap: Useful When You Already Know the Literature

If you have spent substantial time in a field, you may begin by noticing something in the existing evidence that appears missing, inconsistent, limited, or unresolved. Perhaps competing studies produce incompatible results. A theoretical mechanism remains weakly tested. Evidence is concentrated in circumstances that limit how confidently conclusions can be applied elsewhere.

Literature reviews help researchers understand what is already known and can reveal gaps that further research might address. Academic introductions also commonly move through the existing state of knowledge toward gaps, shortcomings, or problems before presenting the study's research question.

The advantage of starting with a gap is that your idea is already positioned relative to existing scholarship. The danger is assuming that absence alone equals significance.

“Nobody has studied X in this exact location” is an observation about the literature. It becomes a stronger justification only when there is a reason that investigating X there could change, test, extend, challenge, or otherwise contribute to what is known.

Topic, Problem, Question, and Gap Are Not Competing Choices

The four concepts are better understood as parts of a developing research argument than as four rival starting methods.

This also explains why distinguishing the research topic, problem, and question is useful even when you do not develop them in a fixed sequence. Each performs a different function in the final logic of the project.

Whatever You Start With, Preliminary Reading Changes the Picture

The literature is not something you consult only after your research question is finished. Preliminary reading helps you discover the terminology used in the field, understand existing explanations and methods, identify relevant debates, and see whether your initial assumptions survive contact with previous research.

Monash recommends beginning investigation of a research topic with reading, including review articles for an overview of the area and recent studies for a picture of current research. Its guidance then moves toward generating a specific question after the researcher has developed an initial understanding of the literature.

How much reading you need before committing depends on the project and how familiar you already are with the area. The important principle is that you should know enough about a possible topic to make an informed decision rather than treating your first impression as established knowledge.

The Process Usually Loops Back on Itself

A neat diagram might show Topic → Gap → Problem → Question. Real projects often behave more like a loop.

You choose a topic, read, and identify what looks like a gap. The gap suggests a problem. You formulate a question. Then the question exposes search terms you had not considered, so you return to the literature. New evidence shows that the supposed gap is smaller than you thought. You revise the problem, narrow the topic, and reformulate the question.

That is not necessarily evidence of poor planning. Iteration is part of making the pieces agree. Even later research writing may require revision as the researcher's understanding develops; Monash notes that thesis introductions are often drafted repeatedly and revised as the project progresses.

What Matters Is Alignment, Not the Origin Story

By the time you have a defensible research plan, a reader should not need to know which idea occurred to you first. They should be able to see the logic of the study.

The topic establishes the area. The existing literature and context establish what is known. The gap or problem helps explain why further investigation is justified. The question states what you will investigate. The methodology then needs to provide a credible way of answering that question.

Research proposals make this alignment visible by presenting the proposed topic, explaining its significance, positioning it in published literature, and setting out an approach for gathering evidence.

Four Starting Points Can Converge on the Same Study

The hypothetical study does not become stronger because one researcher happened to start with the gap while another started with an observation. Its strength depends on whether the eventual problem is supported, the question is meaningful, the existing evidence has been understood accurately, and the proposed study can answer the question credibly.

The example also shows why research ideas can emerge from many different sources. What happens after the initial idea is usually more consequential than which doorway you entered through.

There Is No Magic Starting Point

Start With What You Actually Have

Instead of trying to manufacture whichever starting point a diagram tells you should come first, identify the strongest thing you currently know. Then use it to determine your next task.

Whichever row describes you, expect movement between the pieces. A preliminary question can send you back to the topic. A literature search can overturn your proposed gap. A problem can reveal that the topic is much broader than you realized.

As the project becomes clearer, narrow the research topic in response to what the problem and evidence actually require rather than adding arbitrary restrictions simply to make the project look specific.

Check Whether Your Starting Point Is Developing Into a Study

Questions About Where Research Should Begin

Start Somewhere, Then Make the Pieces Agree

How to Cite This Guide

This guide is intended to be read, shared, and used in research, teaching, and academic work. If you draw on its ideas, explanations, or other content, please acknowledge the source by citing the guide. Doing so gives appropriate credit and helps your readers locate the original resource.