How to write a strong research question

Every thesis hangs off one sentence: the research question. Get it right and the rest of the work has a spine. Get it wrong (too broad, already answered, or impossible to actually study) and you feel the drift for years. The hard part is that a good question has to satisfy four things at once, and you usually can't tell whether it does until you've read the field. Question Builder inverts that: it drafts candidate questions grounded in the literature first, so you choose from real options instead of guessing in the dark.

What makes a research question strong

A defensible question is specific enough to answer, novel enough to be worth answering, feasible for one graduate student in the time you have, and grounded in a body of work you can build on. Those pull against each other: the most novel question is often the least feasible; the most feasible is often already answered. Writing a good question is really about finding the point where all four hold at once.

Two framings help you get there. PICO (Population, Intervention, Comparison, Outcome) forces an empirical question into testable parts. FINER (Feasible, Interesting, Novel, Ethical, Relevant) is a checklist for whether a question is worth committing to. Question Builder frames each candidate as one or the other, so the structure is there before you start editing.

1. Start from an interest, not a finished question

Open Question Builder and describe what you want to study in plain language. 'Something about spaced repetition and exam performance' is enough. You get back candidate questions, each PICO or FINER framed, drawn from the actual literature rather than invented. If the corpus is thin on your topic, the builder sweeps the live literature first, so your candidates are never framed on near-strangers.

You can generate more candidates to widen the pool, sort them, and remove the ones that don't fit. Nothing is committed yet: this is the exploration stage, and it is meant to be cheap to throw away.

2. Read the signals, and know what they are

Each candidate carries three scores: Evidence (how much prior work supports investigating it), Novelty (how under-explored it is), and Feasibility (how achievable it is for one student given data access, scope, and time). Sorting by any of them surfaces a different shape of question: sort by Novelty to find the untrodden ground, by Feasibility to find what you could realistically finish.

One honest distinction the tool is careful about: these three are AI estimates from the evidence, not measurements. They are a fast way to compare candidates, not proof that the literature backs the question. For that, there is a separate, measured check.

3. Run the grounding check before you commit

Select a candidate and the grounding check tells you, from real data, how many sources in your own library and in the wider corpus actually address it, and names the strongest anchor papers (with their open-access status). This is the reality test the scores can't give you: a question can look novel and feasible and still have nothing under it. If the grounding check comes back thin, the fix is one click to Literature Finder to go build the evidence base before you pin.

4. Pin the one you'll defend

When a candidate holds up, pin it as your thesis question. This matters more than it looks. The pinned question becomes the authoritative aim every downstream tool frames its work around: your literature search, your synthesis, your chapter drafts, and your viva rehearsal all point back to it. Pinning is a deliberate, human decision; the tool never chooses your question for you.

Already have your question?

If you arrived with a question already in hand, you don't have to generate anything. Type it in and pin it directly, and it becomes the same authoritative aim every drafting tool builds on. Either way, the next move is the same: build the conceptual framework around it, then go find sources. Read "Build a conceptual framework grounded in the literature" next, or jump straight to the getting-started walkthrough to draft your first grounded section.