A research gap is a claim about the literature: something the field has not established, and a reason that matters. A problem statement is a claim about the world or the discipline that follows from it, narrow enough for one study to address. Postgraduate examiners test both by asking what would change if you were right.
Most guidance stops at that definition. What follows is the opposite: six gap statements and six problem statements written out in full, one per discipline, at the length they occupy in a real Master’s dissertation or PhD thesis, plus a fill-in template and the failures that get named at upgrade. The move-by-move structure of the chapter these paragraphs live in is covered separately in our guide to the doctoral introduction chapter; this piece is about the two paragraphs themselves.
What is the difference between a gap and a problem?
They answer different questions and they fail in different ways. The gap statement answers “what does the field not know?” and is evidenced entirely from the literature; it fails by being unfalsifiable, because nobody can check “no one has studied this”. The problem statement answers “why does that matter, and to whom?” and is evidenced from the world, from policy, from practice or from the internal logic of the discipline; it fails by being too large for one study to touch.
The pairing is what examiners actually assess. A precise gap attached to a vague problem produces a thesis that is technically sound and pointless. A vivid problem attached to a vague gap produces a thesis that reads like journalism. Write them as two paragraphs, in that order, and make the second one narrow the first.
Which kinds of gap survive a viva?
Not all of them are equal, and the weakest is the most common. The table below sets out the five that appear in postgraduate work, what each claims, and what an examiner will ask for.
| Type of gap | What it claims | What the examiner asks for | Strength |
|---|---|---|---|
| Population or setting | The relationship is established, but not in this population, sector or country | A reason the case should behave differently, not merely that it is different | Weak on its own, strong with a stated mechanism |
| Empirical contradiction | Two well-conducted bodies of work disagree and nobody has explained why | Both bodies of work, cited, and a candidate explanation you can test | Strong |
| Methodological | The finding may be an artefact of how the field has measured it | Evidence that the measurement choice is shared across the literature | Strong |
| Theoretical | An existing theory has not been extended to, or tested in, this domain | Why the extension is non-trivial rather than mechanical | Moderate to strong |
| Assumption | The literature takes something for granted that may not hold | Where the assumption is visible in the sources, and what follows if it fails | Strongest, hardest to write |
The ordering is not arbitrary. “Nobody has studied X in Y” invites the reply that nobody needed to; every other row survives that reply. If your only available gap is a population gap, rescue it by naming the mechanism that makes your population informative — a regulatory difference, a funding structure, a demographic composition — and cite the source of that difference.
What does a gap statement actually look like?
Six worked pairs follow. Each gap paragraph is evidenced from literature; each problem paragraph is evidenced from the world. Both are written at the length they occupy in a submitted thesis, and the bracketed prompts show where your own citations go. They are illustrative specimens, not claims about any real literature — replace every bracket with a source you have actually read.
Management
Gap. “Studies of hybrid working have converged on the individual level, measuring productivity and wellbeing outcomes for employees [cite three]. The supervisory layer that has to make hybrid arrangements operate has been studied far less, and the two studies that address it reach opposing conclusions: one reports that middle managers compensate for reduced visibility by increasing output monitoring [cite], the other that they shift towards outcome-based judgement [cite]. Neither accounts for the other, and neither specifies the organisational conditions under which each response occurs. The field therefore lacks an explanation of when reduced co-presence produces tighter control rather than looser control.”
Problem. “The choice between those two responses is consequential for organisations that have made hybrid arrangements permanent, because monitoring and outcome-based judgement generate different costs, different employment-relations risks and different retention effects. Middle managers are making that choice now, without evidence, and the conditions that shape it are unknown. This thesis addresses that by explaining how middle managers in [sector] construct control under hybrid arrangements, and what organisational conditions push them towards each response.”
That pair is an empirical-contradiction gap. It is deliberately not a hypotheses-and-variables piece; how a management doctorate moves from a problem like this to testable hypotheses and operationalised constructs is set out in our guide to research questions, hypotheses and variables for a management doctorate.
Education
Gap. “Research on the attainment effects of [intervention] has been conducted overwhelmingly in primary settings [cite]. Where secondary studies exist they measure attainment at the end of a single key stage [cite], so the literature cannot say whether effects persist, decay or reverse across a transition. More fundamentally, the mechanism is usually left implicit: studies report that the intervention works without specifying what it changes in classroom practice, which means that null results cannot be distinguished from implementation failure.”
Problem. “Schools are being asked to adopt [intervention] on the strength of primary-phase evidence, and are doing so at a point in the curriculum where the mechanism may not operate. If the effect depends on [mechanism], secondary adoption will underperform in a way that looks like a failure of the intervention rather than of the transfer. This thesis addresses that by identifying what changes in classroom practice when [intervention] is adopted at key stage 3, and whether those changes are the ones the theory requires.”
That pair rests on a theoretical framework doing real work; which framework an education doctorate should build on, and how to keep it steady, is covered in our piece on theoretical frameworks for an education PhD or EdD.
Nursing and health
Gap. “The effectiveness of [intervention] has been established in trial conditions [cite two]. Implementation studies report substantial variation in real-world delivery [cite], but attribute it mainly to staffing levels. Qualitative accounts from practitioners point instead to the fit between the intervention protocol and existing ward routines [cite], and this explanation has not been tested against the implementation data. The literature consequently cannot say whether variation in delivery is a resourcing problem or a protocol-design problem.”
Problem. “Those two diagnoses lead to opposite responses: one funds more staff, the other redesigns the protocol. Services are currently choosing between them without evidence, and the cost of choosing wrongly is an intervention that is judged ineffective when it was never delivered as designed. This thesis addresses that by examining how [intervention] is adapted in practice in [setting], and whether the adaptations track staffing or protocol fit.”
Engineering
Gap. “Models of [phenomenon] are validated almost exclusively against laboratory specimens under controlled loading [cite three]. Field data are scarce, and the two published field datasets differ from laboratory conditions in [variable], which the models treat as constant. Whether the models retain their predictive accuracy when that assumption is relaxed has not been established, because no study has validated them against field measurements collected under varying [variable].”
Problem. “Design codes draw on these models to set safety factors, so an unexamined assumption propagates into structures in service. If the models degrade under field conditions, current factors are either unsafe or uneconomically conservative, and there is at present no basis for saying which. This thesis addresses that by validating [model class] against field measurements from [structure type] and quantifying the error introduced by the constant-[variable] assumption.”
Law
Gap. “The doctrinal literature on [provision] has concentrated on its interpretation by the appellate courts [cite]. Commentary assumes that first-instance application follows appellate reasoning, and the assumption is untested: no study has examined how the provision is applied at first instance, where the overwhelming majority of decisions are taken and few are appealed. The scholarship therefore describes a body of law as it is articulated rather than as it operates.”
Problem. “If first-instance application diverges from appellate reasoning, the practical content of the provision differs from its doctrinal statement, which matters for parties who will never reach an appellate court and for any reform argument built on the reported cases alone. This thesis addresses that by analysing first-instance application of [provision] and comparing it with the appellate line.”
A doctrinal thesis of this kind organises itself by legal problem rather than by the social-science chapter sequence, as our account of law PhD thesis structure sets out; and every authority in it has to be cited to the current standard, which changed with the fifth edition of OSCOLA — see our guide to citing UK legislation and case law.
Politics and international relations
Gap. “Explanations of [outcome] divide between accounts emphasising domestic coalitions [cite] and accounts emphasising external constraint [cite]. Both are supported by their own case selections, and the cases rarely overlap. No study has examined a case in which the two explanations predict divergent outcomes, so the literature has not been placed in a position to discriminate between them.”
Problem. “Which explanation holds determines where analysts and policymakers look for leverage over [outcome], and the two point in incompatible directions. This thesis addresses that by selecting [case] precisely because the two explanations diverge there, and tracing which mechanism operated.”
How does a Master’s dissertation calibrate this differently?
The structure is identical; the scope of the claim is not. A doctoral gap claims something about a field and has to be defended against everyone working in it, which is why the contribution has to be statable in one sentence — see our piece on what counts as an original contribution to knowledge. A Master’s gap claims something about a bounded literature: a sector, a setting, a policy period, a specific instrument. “The literature on X in the UK further education sector” is a legitimate Master’s framing and an insufficient doctoral one.
Length differs accordingly. A Master’s gap and problem occupy perhaps 250 words between them; a doctoral pair often runs past 800 because each claim about the literature has to be evidenced. The commonest Master’s error is borrowing doctoral scope and then delivering a study that cannot reach it, which converts a good dissertation into an overclaimed one.
What template can you fill in?
Five slots, in order. Fill each with one sentence before you write anything continuous.
- Established. What the literature has settled, with the two or three sources that settled it.
- Limit. The specific thing those sources do not establish — a contradiction, an assumption, a measurement choice, an untested extension.
- Evidence of the limit. How you know: the studies that disagree, the assumption visible in their design, the measure they share.
- Consequence. What follows in the world or the discipline if the limit is not resolved, and for whom.
- Address. One sentence beginning “This thesis addresses that by …”, naming the study, not the aspiration.
If slot 3 is empty, you have an assertion rather than a gap. If slot 4 is empty, you have a gap with no problem and the thesis will be asked in the viva why it was worth doing. That question is also asked early, at the progress review that confirms your registration — our guide to the MPhil to PhD upgrade covers what the panel is deciding.
Which failures do examiners name?
Five, and they recur. A gap defined by absence alone, with no reason the absence matters. A problem statement that is a description of a social issue rather than a claim your study can address. A mismatch between the two, where the problem is national and the gap is about one instrument. A gap that quietly disappears after the introduction, so that the findings answer a different question from the one the chapter set up. And a gap written at the start of candidature and never revised, which by submission describes a literature that has moved — the newest relevant paper is the one that most often removes a gap, which is why the claim has to be rechecked before you submit, not before you begin.
Holding the claim steady from upgrade to submission
The gap and the problem are the two paragraphs most likely to drift, because they are written early, quoted in funding applications and progress reports, and then left alone while the study changes around them. The discipline that prevents it is mechanical: keep the pair in one place, revisit it whenever a finding lands, and make the introduction’s final draft the last thing you write. Draft your thesis in Tesify and hold the gap statement, the problem statement and every chapter in one workspace, so that the claim you defend in the viva is the claim your findings actually support. The argument stays entirely yours; there is a free plan, and the useful first step is to write the five template slots as five plain sentences.
Frequently asked questions
How long should a research gap statement be?
In a Master’s dissertation, roughly a paragraph of 120 to 150 words, with the problem statement a similar length. In a PhD thesis the pair commonly runs to 600 to 800 words, because every claim about the literature has to be evidenced by named sources rather than asserted.
Is “no one has studied this” a research gap?
It is the weakest form and examiners treat it as one, because the obvious reply is that nobody needed to. Rescue a population or setting gap by naming the mechanism that makes your case informative — a regulatory difference, a funding structure, a demographic composition — and citing the source of that difference.
Where does the gap statement go in the thesis?
It appears twice. A compressed version sits in the introduction, where it establishes why the study exists, and the fully evidenced version closes the literature review, where the synthesis that produced it is on the page. The two must say the same thing.
What is the difference between a problem statement and a research question?
The problem statement is a claim about why something matters and what is unresolved. The research question is what you will actually answer, and it must be narrower than the problem and answerable with the data you can obtain. A question that restates the problem has not narrowed anything.
Can my gap change during the PhD?
It usually does, and examiners expect it to. What they object to is an introduction that still describes the original gap while the findings answer a different one. Rewrite the pair after the findings are stable, and treat the introduction as the last chapter you finalise.
How many sources should support a gap claim?
Enough that the claim is checkable, which in practice means naming the specific studies whose limits you are identifying rather than citing a block of ten. A gap supported by two precisely characterised studies is stronger than one supported by twenty listed in parentheses.
Do qualitative theses need a problem statement?
Yes, and the requirement is identical. What differs is the evidence: a qualitative problem statement is more often grounded in practitioner accounts, policy documents or an unexplained pattern in existing fieldwork, rather than in a numerical trend.
Is a research gap the same as an original contribution?
No. The gap is what the field has not established before you start; the contribution is what your completed study added. A thesis can identify a real gap and fail to contribute, which is why the contribution has to be stated separately and in one sentence.
What if I find a paper that closes my gap?
Read it properly before panicking, because most such papers narrow a gap rather than close it — different population, different measure, different period. If it genuinely closes it, the honest move is to say so in the literature review and reposition your study against it; an examiner who finds that paper themselves, unmentioned, will draw a worse conclusion.