Start writing free

How it works FAQ Blog Help Contact Log in

Category: Thesis Guides

You Have Read 300 Papers and Have No Literature Review Chapter. How to Build One (2026)
Here is the failure, and almost every doctoral researcher writes it once. You have read for two years. You open a document and start working through what you read: a paragraph on Ahmed, a paragraph on Brennan, a paragraph on Chowdhury, each accurately summarising a study. Forty pages later you have something that is thorough, defensible, laborious and not a literature review.

It is an annotated bibliography with the headings removed. The diagnostic takes five seconds: look at how your paragraphs begin. If most of them open with a researcher’s name, you have organised by source. A literature review is organised by claim.

This page is about closing that gap, and it assumes the reading is done. If the problem is that you have notes rather than sources, the earlier stage is covered in our workflow for turning three years of notes into chapters.

This chapter is assessed before your thesis is

Worth knowing early, because it changes how seriously candidates take it. The literature review is not only a thesis chapter; at many institutions it is a formal progression requirement in its own right.

UCL’s criteria for upgrading from MPhil to PhD status include “Demonstration of sufficient awareness of the context of the work and completion of such tasks as a review of relevant literature and a bibliography”, alongside the ability “to formulate a viable hypothesis or research question that could be completed within the normal time frame”. Those two sit next to each other for a reason: the panel is checking that your question comes out of the literature rather than being asserted alongside it. The wider mechanics of that gate are in our guide to the MPhil to PhD upgrade and confirmation review.

So this chapter gets read closely twice by different people with different jobs: once by a panel deciding whether you should continue, and once by examiners deciding whether you have contributed anything. Both are reading for the same thing, which is the subject of the next section.

What the chapter is actually for

Not to prove you have read widely. Your bibliography proves that, and nobody is impressed by volume.

The chapter exists to build the case that your contribution is necessary. Everything in it should be doing one of three jobs: establishing what is known, establishing how it is known, or establishing what is not known. The third is the one your thesis exists to change, and a review that never arrives there has failed regardless of how much it covers.

This is why the review and the contribution claim are the same argument told at different lengths. Your one-sentence statement of your original contribution to knowledge is the compressed version; the literature review is the expanded version that makes it credible. If they do not match, one of them is wrong, and it is usually not the review.

Most of what you read does not go in. Reading it was not wasted — it is what lets you judge the twenty per cent that does.

The organising decision that fixes most of it

There are four ways to order a literature review, and choosing wrongly causes more damage than any amount of poor prose.

By source. One paper per paragraph. Never correct in a doctorate. This is the failure described at the top of this page, and no amount of editing rescues it, because the problem is the skeleton rather than the surface.

Chronologically. Occasionally right, and much more rarely than it is used. Chronology earns its place only when the development of the idea is itself your point — when you are showing how a field changed its mind and why that matters to your question. Used by default, it produces a narrative with no argument, because “and then someone else did a study” is not a claim.

By theme or construct. The default for most doctorates, and correct most of the time. Each section handles one idea and reports what the field collectively knows about it, with the individual studies cited as evidence within your sentence rather than as the subject of it.

By debate. The strongest option where your field genuinely disagrees. Each section states a contested question, sets out the positions and the evidence for each, and explains what would settle it. This structure is naturally hospitable to a gap, because unresolved debates are gaps.

The practical difference is visible at sentence level. “Brennan (2021) found that supervisory contact predicted completion” is organised by source. “Supervisory contact predicts completion in several settings (Brennan, 2021; Chowdhury, 2023), though both studies aggregate across disciplines that differ sharply in independence norms” is organised by claim — it states what is known, cites who established it, and begins the move toward what is not known.

The six moves

A sequence that works whether you have 60 sources or 400.
1. Write the gap sentence first. Before drafting a word of the chapter, write down what your thesis will establish that is not currently established. It will change; write it anyway. It is the destination, and a review written without one wanders.
2. Sort your sources into claims, not piles. Go through the library and, for each source, record the one thing it establishes that matters to your question — not what it is about, what it shows. Sources that resist this are usually sources you do not need.
3. Cluster the claims. Group them until you have somewhere between four and eight clusters. Fewer than four usually means the clusters are too coarse to say anything; more than eight usually means you are still organised by topic rather than by argument. These clusters are your sections.
4. Order the clusters so each one earns the next. Section two should be a question raised by section one. If the sections could be shuffled without loss, you have a list rather than an argument, and the fix is at this stage rather than in the writing.
5. Draft each section to a fixed shape. What the field holds, on what evidence, with what limitation, and what that leaves open. Four moves per section, in that order, every time. It feels mechanical while drafting and reads as authoritative afterwards.
6. Write the closing synthesis last. Two or three pages that collect the “left open” sentences from every section and converge them on your research question. This is the passage your upgrade panel and your examiners will read most carefully, and it should make your question feel inevitable rather than merely reasonable.
Step three is the whole job. Once the clusters are right, the drafting is transcription; while they are wrong, no amount of rewriting helps.

Five failures examiners name

The review that never comes back. The commonest structural criticism of a thesis: a competent literature chapter that is then never used again, so the findings are discussed against nothing. Every cluster you build should reappear in the discussion chapter. If a section has no counterpart there, ask what it is doing in the thesis.

The gap that is merely an absence. “No one has studied this in Wales” is an absence, not a gap. A gap requires a reason the missing knowledge matters: something about the Welsh context that should make the established relationship behave differently. Absences are infinite and uninteresting; gaps are arguments.

Uncritical coverage. Reporting what studies found without ever assessing how well they found it. You are being examined on judgement, so a review in which no study has a weakness reads as one where none was read closely.

Straw-manning the field. Overstating what is missing in order to make your contribution look larger. Examiners know this literature better than you do, and an overstated gap is the easiest thing in a thesis to puncture. Claim precisely what you can defend.

Staleness. A review completed in year two and never revisited. Re-run your searches before submission, and check that nothing load-bearing has been withdrawn in the meantime — a step covered in our guide to auditing a bibliography against the retraction record.

How long, and how many sources?

There is no rule, and the honest answer is that both follow from the clusters. In a monograph-shaped thesis a single literature chapter commonly runs somewhere around 8,000 to 12,000 words, in line with other substantive chapters; where two genuinely distinct bodies of work are in play, two chapters are normal rather than indulgent. What decides it is your argument rather than a target, and the reasoning behind splitting or merging chapters at all is set out in our answer to how many chapters a PhD thesis has.

On source count, the useful test is not a number but coverage of the clusters: every claim you make should rest on enough evidence that an examiner in the field would not immediately name something you have missed. That is a far more demanding standard than “300 references”, and a far more achievable one than “everything”.

The mechanical layer, which is not optional

Two infrastructure decisions determine whether this chapter is difficult or merely long.

Your references must be in a manager, with DOIs, before you draft. A literature chapter is reordered more than any other part of a thesis — clusters merge, sections swap, whole passages move to the discussion — and every one of those moves is a citation operation. Doing it by hand is how bibliographies drift out of step with the text, which is the single most avoidable examiner criticism there is. Which manager to use, and what it costs at doctoral scale, is covered in our comparison of Zotero, Mendeley and EndNote.

Finding is a different job from synthesising. If the clusters keep revealing holes, that is a discovery problem rather than a writing problem, and the tools for it are compared in our roundup of AI research assistants for doctoral work. Do not try to solve it inside the drafting session; you will produce neither a search nor a chapter.

What remains after that is the actual work: holding four to eight clusters, several hundred sources and one argument in view at once across a chapter you will draft over weeks. Tesify holds that structure section by section and formats your references as you cite them, so the chapter stays coherent while it is being reordered and the bibliography never falls behind the text. The reading, the clustering and every sentence of the argument stay 100% written by you — which is the only version you can defend when a panel asks why your question follows from this literature. It is free to start.

When to write it

Draft it early for the upgrade, then rewrite it late. These are genuinely two different chapters wearing the same title. The upgrade version argues that a viable question exists; the final version argues that your completed work answers it, and it is written with the knowledge of what you actually found.

Candidates who treat the upgrade draft as finished spend the final year fighting a chapter that no longer matches the thesis. Expect to rewrite the closing synthesis completely, and expect at least one cluster to turn out to matter far more than you thought. Rewrite it before the introduction, since the introduction has to promise what this chapter sets up.

Frequently asked questions

What is the difference between a literature review and a summary of the literature?

Organisation. A summary is arranged by source and reports what each study said. A review is arranged by claim and reports what the field collectively knows, citing studies as evidence inside the sentence rather than as its subject. If your paragraphs begin with authors’ names, you have written the first.

How should I structure a PhD literature review?

By theme or construct in most cases, by debate where your field genuinely disagrees, and chronologically only when the development of the idea is itself your argument. Never by source. Aim for four to eight sections, ordered so that each raises the question the next answers.

How long should a doctoral literature review chapter be?

Commonly around 8,000 to 12,000 words in a monograph-shaped thesis, in line with other substantive chapters. Two chapters are normal where two distinct bodies of literature are in play. The clusters should decide the length rather than the reverse.

How many sources does a PhD literature review need?

Enough that each claim you make is properly supported and an examiner in your field would not immediately name something obvious you have missed. That is a coverage test, not a count, and it is both stricter and more achievable than any number.

Is the literature review required for my upgrade?

Frequently, yes. UCL’s upgrade criteria explicitly include “completion of such tasks as a review of relevant literature and a bibliography” alongside the ability to formulate a viable research question. Check your own criteria, since the requirement and its form are institutional.

What counts as a research gap?

A missing piece of knowledge plus a reason it matters. “Nobody has studied this population” is an absence; “this population differs in a way that should change the established relationship, and nobody has tested it” is a gap. Examiners can tell the two apart instantly.

Should I criticise the studies I review?

Yes, proportionately and specifically. You are examined on judgement, and a review where no study has a limitation reads as one where nothing was read closely. Criticise the design and the inference, not the researcher.

Do I have to include everything I have read?

No, and trying to is the main cause of overlong, shapeless chapters. Include what your argument needs. The rest of your reading is not wasted; it is what allows you to judge which sources matter.

Should the literature review come before or after the methodology?

Before, in the conventional monograph shape, because the method is justified by the gap the review establishes. In some by-publication and humanities theses the literature is woven through instead, which is a disciplinary convention worth checking against recent theses from your own department.

When should I update my literature review?

At minimum once between the upgrade draft and submission, and again shortly before you submit. Re-run your searches, add anything significant, and check that nothing you rely on has been retracted while you were writing.

Can I reuse my upgrade literature review in the final thesis?

As a foundation, yes; as a finished chapter, rarely. The upgrade version argues that a good question exists, while the final version argues that your completed work answers it, so the closing synthesis in particular usually needs rewriting from scratch.

What is the most common examiner criticism of a literature review?

That it is never used again. A strong review whose themes do not reappear in the discussion leaves the findings interpreted against nothing, and it is the structural fault examiners raise most often.
August 23, 2026
Preparing for Your Viva with AI: Rehearsing the Questions Examiners Will Actually Ask
Your thesis is submitted, the date is set, and you are about to spend two to four hours defending three or four years of work to two people who have read it more recently than you have. The instinct is to reread the thesis. That is necessary but insufficient — the viva is an oral examination, and reading silently does not rehearse speaking.

What follows is a rehearsal method rather than a reading list, and it is built from the one asset you already have: your own thesis. If you want help turning it into a question bank you can practise against, you can do that in Tesify.

What the viva is for

Two purposes, and understanding both changes how you prepare.

The first is to establish that the thesis makes an original contribution. The second, less discussed, is to verify that the work is yours. The QAA notes that in the UK examiners do not normally reveal the outcome to the candidate in advance of the oral examination, precisely because one of its purposes is to check both the authorship of the thesis and the candidate’s engagement with the described research process.

That second purpose is why the questions are so often about why rather than what. An examiner establishing authorship does not ask you to recite your findings. They ask why you chose this method over the obvious alternative — because only someone who made the decision can answer.

The eight questions almost every candidate is asked

Prepare an answer to each of these, out loud, before you prepare anything else.
1. Summarise your thesis in a few minutes. Almost universal as an opener. Have a two-minute version and a five-minute version.
2. What is your original contribution? One sentence, then elaboration if invited.
3. Why did you choose this topic? Sounds like small talk. It is establishing that the project was yours from the start.
4. Why this methodology and not [obvious alternative]? Expect this for every major design decision.
5. What are the limitations of your study? Volunteer three, each with its consequence analysed.
6. What would you do differently? Answer substantively. “Nothing” reads as a lack of critical distance.
7. How does your work relate to [named study]? Usually a paper by the external examiner or someone in their circle.
8. What next — publication, further research? Have a specific answer, not an aspiration.
If you can answer these eight fluently and aloud, you have covered a substantial share of most vivas.

Generating the questions only your thesis provokes

The generic list gets you to competent. What distinguishes a comfortable viva is having anticipated the questions specific to your thesis — and here AI assistance is genuinely well suited, because generating challenges to an argument is a structural task rather than an interpretive one.

Work through your thesis chapter by chapter and, for each, produce questions of these kinds:
- Decision questions. Every methodological choice invites “why not otherwise?”. Sampling, instrument, analytical approach, exclusions, time period, case selection.
- Weak-point questions. Where is the argument thinnest? A small sample, a contested measure, a chapter that leans on one source, an inference the data only partly supports.
- Boundary questions. How far do your claims generalise? Where would they stop holding?
- Contradiction questions. Where does your work disagree with established findings, and how do you account for it?
- Definition questions. Every key construct: how did you define it, and why that definition?
Feed your own chapter summaries, methodological decisions and known limitations in, and ask for the hardest questions an examiner could put. You are not asking a tool to assess your work — you are asking it to stress-test a structure you built. The answers must be yours.

Two chapters deserve the most attention. Your methodology, because that is where decisions cluster. And your discussion, because that is where the contribution is argued — the moves it makes are set out in our guide to writing the discussion chapter, and examiners test each one.

A question bank is only useful if you answer from it aloud — silent rehearsal does not transfer to the room.

The step candidates skip

Answer out loud. Not in your head — aloud, in full sentences, ideally to another person or a recording.

The gap between knowing something and being able to explain it under mild pressure is large, and it only closes with practice. Doctoral researchers who have rehearsed silently routinely discover in the room that a point they understand perfectly comes out as three false starts.

Record yourself answering the eight standard questions and listen back once. It is uncomfortable and it is the highest-yield hour of preparation available. Listen for whether you answered the question asked, whether you stopped when you were done, and whether you hedged so heavily that the claim disappeared.

Then arrange a mock viva if your department offers one. A mock with people who have not read your thesis closely is still worth doing — much of what you are practising is the act of explaining, not the specific content.

Know what can actually happen

Anxiety feeds on vagueness, so replace it with the published facts. Examiners at UK institutions have a defined set of recommendations available to them, and corrections are the ordinary outcome rather than a judgement against you — our data on UK viva outcomes across 127 institutions sets out exactly which outcomes exist and the timescales attached to each. The QAA’s own position is that only a minority of candidates fail the doctoral assessment outright.

Read your institution’s research degree regulations before the day so you know which outcomes are available to your examiners, and be ready for the contribution question specifically, since it is the one the whole examination circles — our guide to what counts as an original contribution to knowledge covers how to state it in a form that invites questions you have prepared for.

Practical preparation in the final fortnight
1. Reread your thesis with fresh eyes, marking anything you would now express differently. Keep a list — volunteering “on reflection I would frame this differently” is a strength, not an admission.
2. Tab the physical copy so you can find any chapter, table or figure in seconds. Fumbling costs composure.
3. Read your examiners’ relevant work. Not everything — the two or three papers closest to your topic. Question 7 comes from here.
4. Prepare a one-page summary of your contribution, methods and key findings to glance at during the examination if permitted.
5. Prepare your errata list. Bringing a list of typographical errors you have already spotted is a mark of professionalism.
On the day

Take your time answering. A three-second pause reads as considered; a rushed answer to a question you misheard is much harder to recover from. Ask for clarification if you need it — that is normal academic conversation, not weakness.

When you do not know, say so, and then say what you would do to find out. Examiners are not testing recall; they are testing judgement. “I did not examine that, but the way to test it would be…” is a strong answer.

And treat disagreement as the point rather than a threat. A viva is the first conversation you will have with senior colleagues as a peer, and examiners frequently push on a point precisely because it is interesting. Defend your position with evidence, and concede where the evidence genuinely does not support you — being able to distinguish the two is a doctoral skill being assessed.

Where Tesify fits

Tesify works from your own thesis: it helps you build a question bank chapter by chapter, tighten your contribution statement, and structure your answers so you can rehearse against them. It does not answer for you, and the defence you give remains 100% written and spoken by you.

If your corrections land afterwards, the same workflow carries into revising chapters against examiner comments — the drafting approach described in our guide to turning notes into chapters applies equally to turning examiner reports into revisions.

Prepare for your viva in Tesify.

Frequently asked questions

How long does a UK viva last?

Commonly two to four hours, though there is no fixed duration and practice varies by institution and discipline. Length is not a signal of how it is going — a long viva often means the examiners find the work interesting.

Will the examiners tell me the outcome beforehand?

Not normally. The QAA notes that UK examiners do not usually reveal the outcome in advance of the oral examination, since part of its purpose is to verify authorship and your engagement with the research process.

Can I take my thesis and notes into the viva?

Almost always yes for an annotated copy of your thesis, and usually for brief notes, but confirm with your institution beforehand. Tab your copy so you can navigate it quickly.

Is it acceptable to use AI to prepare for a viva?

Preparing is not being examined — the examination is what you say in the room. Using a tool to generate practice questions from your own thesis is a rehearsal aid, much like a mock viva. What matters is that the answers are your own understanding, which is exactly what the examination tests.

What if I am asked about a paper I have not read?

Say so honestly and engage with the substance: ask what it argues, and respond to that. Pretending to know a paper is far more damaging than not knowing it, and examiners sometimes raise very recent work precisely to see how you think on your feet.

Should I have a mock viva?

Yes, if it is available. The benefit is mostly in practising the act of explaining under mild pressure, which is why even a mock with someone outside your specialism has value.

What if I disagree with an examiner?

Say so, with reasons and evidence. A defensible disagreement demonstrates exactly the independent judgement a doctorate certifies. What weakens your position is either capitulating on a point you have good evidence for, or defending one you do not.

How should I handle a question about a limitation I already know about?

Welcome it. Name the limitation, state its consequence for your findings, and explain what you did to mitigate it or how you accounted for it in your conclusions. A limitation you have already analysed is a prepared answer, not an attack.
August 17, 2026
Three Years of Notes and No Chapter: An Honest AI Workflow for Doctoral Writing Up

You have the data. You have three years of reading. You have folders of notes, half-written sections, a supervisor waiting on chapter four and a funded period that ends before the thesis does. What you do not have is a chapter — and every week that passes makes starting harder, because the material keeps growing while the blank document does not.

This is the most common failure point in the UK doctorate, and it is almost never a knowledge problem. It is a problem of converting accumulated material into linear argument. If you want to start that conversion now, you can do it in Tesify.

What writing up actually costs you

Be clear about the stakes, because they are not only academic. An overrun typically means an unfunded writing-up year, during which the stipend has stopped and the work has not. Advance HE’s Postgraduate Research Experience Survey 2025 — 35,475 responses across 93 institutions — found that among doctoral researchers who had considered leaving, 20% cited mental or emotional health and 14% cited financial difficulties. The writing-up period is where those two pressures meet.

And the delay is self-reinforcing. Every month you do not draft, the volume of material to synthesise grows, the memory of why you made a decision fades, and the perceived size of the task increases. The single most valuable thing you can do is convert something — anything — into prose this week.

Why the notes will not become a chapter on their own

Because notes and chapters are different objects. Notes are associative, indexed by when you encountered something. A chapter is linear and argumentative, indexed by what the reader needs next. There is no amount of reorganising your notes that turns one into the other; the conversion is an act of authorship.

The specific trap is waiting to feel ready. Doctoral researchers routinely believe that one more paper, one more analysis, one more pass through the data will make the chapter writable. It will not, because the missing element is not information — it is a decision about what you are claiming.

The staged workflow

Stage 1: Write the claim before the chapter

One sentence: what does this chapter establish? Not what it is about — what it establishes. “This chapter examines participants’ experiences of supervision” is a topic. “This chapter shows that researchers interpret supervisory availability through disciplinary norms about independence, which explains why identical contact hours produce opposite satisfaction judgements” is a claim.

Everything downstream is easier once this sentence exists, because it tells you what to include and, more importantly, what to leave out. If you cannot write it yet, that is the work — not the drafting.

Stage 2: Build a skeleton of claims, not topics

List the five to eight things the chapter must establish, in the order a reader needs them. Each becomes a section, and each section heading should be a claim you could defend, even if the final version is a conventional noun phrase.

This is where AI assistance is genuinely useful and entirely defensible: give it your own claim sentence and your own list of sections and ask whether the sequence has gaps, whether any step assumes something not yet established, and where a reader would object. That is structural critique of your material, and it is the fastest route out of a stall.

Stage 3: Attach your evidence to each claim

Under each section, list the specific evidence that supports it: which participants, which results, which sources. This is where three years of notes finally earn their keep, and it is a sorting task rather than a writing task — which is why it is achievable on a bad day.

If a section has no evidence beneath it, you have found something important. Either the claim is unsupported and must go, or the analysis you need has not been done yet. Better to discover that now than in your viva.

Stage 4: Draft badly, deliberately

Write the worst acceptable version of each section, quickly, without stopping to check citations or polish sentences. The purpose is to get the argument onto the page where you can see whether it holds.

Most doctoral researchers cannot do this, because years of training have made them unable to write a sentence they know is imperfect. This is precisely why writing up stalls: the standard that produces good final chapters produces no first drafts at all. Separate the two activities and both get easier.

Sorting evidence under claims is achievable on a day when writing is not — and it is most of the work.

Stage 5: Revise against the claim

Now bring your standards back. For each paragraph, ask whether it advances the chapter’s claim. If it does not, it goes — however good it is and however long it took. Material cut here can live in an appendix or a future paper.

Stage 6: Verify everything

Every citation opened and checked. Every number traced back to your analysis. Every quotation matched against the source. Do this as a separate pass, because verification and composition use different attention, and doing them together is how errors survive.

The line you cannot cross

This audience will not forgive a fudge on this, so here it is plainly.

What AI can defensibly do: critique the structure of an argument you have built; suggest an order for sections you have specified; tighten sentences you have written; identify where a draft assumes something it has not established; help you get an imperfect first version onto the page so that revision has something to work on.

What it must never do: decide what your findings mean; generate citations or reference lists; summarise papers you have not read; write the contribution claim; or produce text you could not explain and defend.

The test is your viva. UK examiners are checking, among other things, that the thesis is yours and that you engaged with the research process — the QAA notes this is part of why the outcome is not revealed in advance of the oral examination. A chapter you cannot discuss is a chapter that will be found, and being unable to explain your own argument is a far worse position than a rough draft.

Two practical rules follow. Never accept a reference from any tool you have not personally opened and read — fabricated citations are discovered in seconds and are catastrophic at doctoral level. And check your institution’s policy on generative AI and any declaration requirement, since these differ between universities and sometimes between departments.

Where Tesify fits

Tesify is built for stages 2 to 4 specifically — the conversion of material you already own into structured drafted chapters. You bring your claim, your evidence and your reading; it helps you get from a skeleton to prose without the week-long stall in front of a blank document.

It does not read your sources for you, decide your contribution or invent your references. The scholarship, the interpretation and the argument remain 100% written by you, which is the only version that survives examination. More than 9,000 students have used it to get stalled writing moving, across more than 15,000 chapters.

Start drafting your chapter in Tesify.

What to write first

Not the introduction. It has to promise what the thesis delivers, and you do not yet know precisely what that is.

Start with a results or findings chapter, because the material is most concrete and the claims are closest to your data. Then the discussion, which is where the thesis makes its case — our guide to writing the discussion chapter sets out the seven moves it needs. Drafting the discussion usually clarifies your contribution, which is the claim examined in our guide to what counts as an original contribution to knowledge. Write the introduction last.

Two practical points while you are at it. Set your timeline against your actual funded period rather than an assumed one, since as our data on UK PhD stipends and funding shows, UKRI publishes no sector-wide duration. And if your document is going to reach 80,000 words with several hundred references, make sure your writing environment can carry it — our comparison of LaTeX and Word at thesis scale covers where each one fails.

Frequently asked questions

Is using AI to help write my thesis research misconduct?

It depends on your institution’s policy and on what you use it for. Structural critique and sentence-level editing of your own writing are treated very differently from generating content or sources. Read your regulations, follow any declaration requirement, and never submit text you cannot defend.

Will my examiners be able to tell?

The reliable signal is not stylistic — it is whether you can discuss your own reasoning in the viva. Examiners probe why you made particular analytical decisions and what alternatives you rejected. A workflow where you make every interpretive decision yourself leaves you able to answer; one where you did not, does not.

How should I declare AI use in my thesis?

Specifically, where your institution requires it. State what the tool did and what it did not — for example, that it was used for structural feedback and clarity editing on text you wrote, and not to identify sources, summarise literature or generate content. A precise declaration reads as command of the boundary.

What about the confidentiality of my unpublished research?

A fair question, and one you should ask of any tool before uploading a chapter. Unpublished doctoral research may be commercially sensitive, may involve participant data with ethical restrictions, and may affect later publication. Read the provider’s current documentation on storage, access and training use, and check whether your data protection obligations to participants permit uploading their data at all.

Is it worth paying for a writing tool on a stipend?

That is your judgement to make against your own budget, and there is a free tier to test the fit before committing. The comparison worth making is not tool cost against zero, but against the cost of an unfunded overrun term.

I have been stuck for months. Where do I actually start?

With stage 1, on the chapter whose material is most complete. Write the single sentence stating what that chapter establishes. If you cannot, the blockage is analytical rather than a writing problem, and that is the conversation to have with your supervisor this week rather than next term.

Should I tell my supervisor I am using an AI tool?

Yes. Supervisors would far rather discuss your workflow openly than discover it at examination, and many will have a view on what your department expects. Raising it yourself also puts you on the right side of any declaration requirement.

August 17, 2026
How to Write a PhD Research Proposal for a UK Studentship Application (2026)
A studentship proposal is not a miniature thesis. It is a document that has to persuade a panel — usually reading dozens of applications in one sitting, several of them outside their own specialism — that your project is important, original and, above all, deliverable in the funded period by you.

That third criterion is where most rejected proposals fall down, and it is the one applicants think about least. What follows is the proposal written as the panel reads it.

Know the funding landscape before you write

Understanding what a studentship actually consists of shapes what you can credibly propose. UK Research and Innovation sets minimum rates for the studentships funded through its training grants. As published by UKRI, the minimum stipend is £20,780 for 2025 to September 2026, rising to £21,805 from 1 October 2026. Where a London weighting applies, the figures are £22,780 and £23,805 respectively — though UKRI notes that the Engineering and Physical Sciences Research Council does not offer a London weighting, and that grant holders may offer higher stipends within their budget regardless of location. The minimum fee for academic year 2026 to 2027 is £5,238, drawn by your research organisation from the UKRI grant.

One point deserves emphasis because it is widely misreported. UKRI does not publish a single sector-wide award duration. Its terms and conditions devolve that decision, stating that research organisations can decide the duration of a studentship provided it accords with the awarding council’s specifications. So do not write “the standard 3.5-year UKRI studentship” in your proposal. Find the duration specified in the advert you are applying to, and plan to that.

These are minimums, not the whole picture. Individual councils, doctoral training partnerships and centres for doctoral training add their own requirements, and some offer enhanced stipends in specific disciplines. Read the actual advert.

What the panel is assessing

Panels vary, but four judgements recur across UK studentship competitions:
1. Is the question worth answering? Significance to the field, and where relevant beyond it.
2. Is it original? Not merely unstudied — meaningfully new.
3. Is it feasible? Achievable in the funded period, with the resources and access available.
4. Is this the right candidate and the right supervisor? Fit between you, the project and the department.
Structure your proposal so a reader can find the answer to each within seconds. Panels skim before they read.

Section by section

Title and summary

The title should name the topic and signal the approach. Save the clever phrasing; clarity converts better with a panel reading its twentieth application.

The summary is the most consequential paragraph you will write, because for some panel members it is the only part they read closely. In roughly 150 words: the problem, the gap, your question, your approach, and what will be known at the end that is not known now.

Background and rationale

Not a literature review — an argument that builds a gap. Move from the broad field, to the specific debate, to the precise thing that is unresolved, in about a page. Every source you cite should be doing work towards establishing that gap.

The failure mode is a comprehensive survey demonstrating that you have read widely. Panels do not doubt that. They want to see that you can identify the one question that matters.

Research questions or objectives

Two to four, numbered, each answerable. If a question cannot be answered by any study you could realistically run, it is an aspiration rather than an objective.

Make them build. A strong set typically moves from establishing something, to explaining it, to testing the explanation — which also gives your thesis its chapter structure and demonstrates that you have thought about the project as a whole.

Methodology

Specific enough to be credible, brief enough to leave room for the rest. Name the design, the data sources or participants, the sampling or selection approach, and the analytical method. Tie each element back to a numbered research question.

Flag anything that requires permission — ethical approval, data access agreements, fieldwork permits, archive access — and say how you will obtain it. A panel that spots an unaddressed access dependency will assume the project may stall.

A timeline that reserves real time for writing and for ethical approval reads as experience; one that does not reads as optimism.

Timeline

Panels read the timeline as a proxy for judgement. Three things distinguish a credible one.

It reserves genuine time for writing up. A plan where data collection runs to the final months and writing appears as a short closing block tells an experienced panel that you have not done this before.

It includes approvals as scheduled activities. Ethical review takes weeks and sometimes months. Building it into the plan demonstrates you know the process exists.

It contains slack. Fieldwork overruns, recruitment is slower than expected, an archive closes. A plan with no contingency is not a plan.

Map the phases against the duration specified in the advert, and mark your institution’s progression milestones — the upgrade or confirmation review in particular — since panels like to see that you understand the structure of a UK doctorate.

Contribution and significance

State what will be known at the end that is not known now, and who will care. Be precise about the form of the contribution rather than gesturing at importance. Where the funder emphasises impact beyond academia, address it concretely — naming the sector or practice community that could use the finding — rather than with generic assertions.

Fit with the supervisor and department

Name your proposed supervisor and say specifically why their expertise suits this project, referring to their work. Name the departmental resources the project needs: a laboratory, a data centre, a research group, a specialist archive.

Contact your prospective supervisor before submitting wherever the process permits it. Supervisors routinely help shape proposals into fundable ones, and a proposal arriving unannounced is at a real disadvantage against one developed in consultation.

The three most common reasons proposals fail

Too ambitious. The single most frequent flaw. A project that would take a well-resourced team five years, proposed as an individual doctorate. Panels read overreach as inexperience. Narrow the question and say explicitly what you have chosen not to do.

The gap is not real. “No one has studied X in [country]” is not sufficient unless you can say why studying it there would change what we know. Absence is not the same as significance.

The methods do not answer the questions. Surprisingly common — a proposal asking why something happens, then proposing a design that can only establish whether it happens. Check each question against its method explicitly.

Before you submit

Read it as a panel member would: skim the summary, glance at the questions, check the timeline, then decide whether to read properly. If that pass does not tell you what the project is and why it matters, restructure.

Then have someone outside your field read it. If they cannot explain your project back to you, the proposal is doing too much work for a specialist reader and not enough for the generalist who may be assessing it.

Finally, be clear-eyed about the argument you are being asked to make. It is the same argument you will make throughout the doctorate: that this work adds something the field does not already have. Our guide to writing the discussion chapter covers how that claim is eventually cashed out at the other end of the project — and reading it now is useful, because a proposal written with the final contribution statement in mind tends to be sharper.

If the thinking is settled and the drafting is the bottleneck, you can build the proposal section by section in Tesify, working from your own reading and your own design. The argument stays 100% written by you.

Frequently asked questions

How long should a PhD research proposal be?

Whatever the advert specifies — usually somewhere between 1,000 and 3,000 words for UK studentship applications. Exceeding a stated limit is a poor first impression, and some panels stop reading at the limit.

Do I have to stick to my proposal once I start?

No. Doctoral projects evolve, and a thesis identical to its proposal three years on would be unusual. The proposal demonstrates that you can design a feasible project and think like a researcher; it is not a contract.

Should I contact a supervisor before applying?

Wherever the process allows it, yes. Supervisors help shape proposals, confirm feasibility and flag resource issues you could not know about. Some structured programmes advertise pre-defined projects and discourage prior contact — follow the advert.

What is the difference between applying to an advertised project and proposing my own?

For an advertised project, your proposal shows how you would approach a question someone else has set, and the panel is largely assessing you. For a self-proposed project, you are assessed on the project as well, which raises the bar on originality and feasibility.

Can I apply for the same project at several universities?

Yes, and many applicants do. Tailor each one — the supervisor fit and departmental resources sections must be specific to that institution, and a proposal that could have been sent anywhere reads as exactly that.

What stipend will I actually receive?

At least the UKRI minimum where your studentship is UKRI-funded, which is £21,805 from 1 October 2026, or £23,805 where a London weighting applies. Some funders and departments pay above the minimum. Note that non-UKRI studentships set their own rates entirely, so check the advert rather than assuming.

Does the proposal need a budget?

Usually not for a studentship application, since the funding package is fixed. If your project has unusual costs — fieldwork travel, specialist equipment, transcription, participant payments — mention them in the methods or feasibility section so the panel knows you have thought about them.
August 14, 2026
How to Write the Discussion Chapter of a PhD Thesis (with a Worked Example)
The discussion is the chapter examiners read most carefully, because it is where the thesis stops reporting and starts arguing. Your results chapter established what you found. The discussion has to establish what it means, why it matters, and what the field should do differently now — and it has to do that without repeating your results or overclaiming beyond them.

Most struggling discussion chapters fail in one of two directions. Either they restate the findings in slightly different words, producing a second results chapter with more hedging. Or they leap straight to grand claims the data cannot bear. What follows is a structure that avoids both.

What the chapter has to achieve

The QAA’s Characteristics Statement: Doctoral Degree sets out the standard the whole thesis is working towards: all UK doctorates require the main focus of the candidate’s work to demonstrate an original contribution to knowledge in their subject, field or profession, through original research or the original application of existing knowledge or understanding.

Your discussion is where that contribution becomes explicit. Everything before it is evidence. This chapter is the argument the evidence supports.

The seven moves

Move 1: Restate the contribution in one paragraph

Open by answering your research questions directly. Not the methods, not the caveats — the answers. Examiners should be able to read your first paragraph and know what the thesis claims.

A worked example:

This thesis set out to establish whether X mediates the relationship between Y and Z in [context], and, if so, under what conditions. It found that X accounts for a substantial proportion of that relationship in [setting A] but not in [setting B], and that the difference is attributable to [mechanism]. The contribution is twofold: it identifies a boundary condition on a relationship that the existing literature has treated as general, and it provides the first evidence from [context], which prior work has not examined.

That is roughly a hundred words and it does more work than most opening pages. Note the structure: what you asked, what you found, what is new about it.

Move 2: Interpret each key finding against the literature

Take your findings one at a time. For each, do three things: state it plainly, place it against what the literature predicted, and account for the relationship between the two.

The account is the part that carries credit. If your finding agrees with prior work, say what it adds — replication in a new context, greater precision, a mechanism the earlier work inferred but did not test. Agreement is not automatically a contribution; you have to say why yours is not merely another instance.

If your finding disagrees, this is your most valuable material, and it needs to be handled with care rather than triumph. Work through the candidate explanations in order: differences in sample or context, differences in measurement, differences in analytical approach, and only then a substantive theoretical difference. Examiners are impressed by candidates who consider that the discrepancy might be methodological before concluding that the field was wrong.

Move 3: Build the findings into a single argument

This is the move most commonly missed. A discussion organised strictly finding-by-finding reads as a list. Somewhere in the chapter, the separate results have to be brought together into a claim that none of them makes alone.

Ask yourself: what does the whole thesis say that no individual chapter says? That sentence is your thesis in the older sense of the word, and it belongs here explicitly.

The discussion is usually the most heavily revised chapter in the thesis — expect several passes before the argument holds.

Move 4: State the theoretical contribution precisely

Vague contribution claims are the most common weakness in doctoral discussion chapters. “This study contributes to the literature on X” says nothing.

Be specific about the form the contribution takes. Useful shapes include: extending a theory to a population or setting where it had not been tested; identifying a boundary condition where a relationship does not hold; reconciling two literatures that disagreed; supplying a mechanism where prior work established only an association; or developing a construct or instrument that did not previously exist.

Name which one you are claiming. A precise, modest contribution is far stronger than a sweeping vague one, and it is much easier to defend under examination.

Move 5: Draw out practical or policy implications, carefully

If your work has applied implications, state them — but keep them proportionate to your evidence. The standard failure is a cross-sectional study that ends with confident recommendations for practice.

The discipline here is simple: for each implication, be able to point at the specific finding that licenses it. If you cannot, it is speculation, and it should be labelled as a direction for future work rather than a recommendation.

Move 6: Write limitations that strengthen the thesis

Counter-intuitively, this section can add to your credibility rather than subtract from it. Examiners will find the limitations regardless. The question is whether you found them first.

Write each limitation in three parts: what the constraint was, what effect it plausibly had on your findings, and what you did or would do to mitigate it. A limitation stated without its consequence is not analysis — it is a disclaimer.

Compare these. “The sample was drawn from a single institution, which is a limitation.” Versus: “The sample was drawn from a single institution, which means the findings may reflect features of that institutional context — in particular its unusually high staff-to-researcher ratio — rather than the sector generally. The direction of any resulting bias would be to overstate the effect of supervisory contact, so the estimate reported here should be treated as an upper bound.”

The second demonstrates command of your own study. Avoid, though, the opposite failure of listing every conceivable weakness until the thesis appears to establish nothing. Three or four substantive limitations, properly analysed, are worth more than a dozen perfunctory ones.

Move 7: Point forward with specific future work

“Further research is needed” is filler. Specify the study you would run next, and say what it would resolve that yours could not. If your work generated a hypothesis you could not test, name it. Examiners frequently ask what you would do with another three years, and a discussion that has already answered puts you ahead.

The four questions examiners bring to this chapter
1. What is new here? Answerable in one sentence, or the contribution is not clear enough.
2. Do the data actually support this claim? Every claim traceable to a specific result.
3. Have they engaged with work that disagrees? A discussion citing only supportive literature reads as selective.
4. Do they know the weaknesses of their own study? Answered by the quality of the limitations section.
How long, and how much overlap with the conclusion?

Length varies by discipline and by whether your thesis has a separate concluding chapter. Where the two are separate, the working division is that the discussion interprets findings against the literature, while the conclusion states the contribution, the implications and the closing argument without introducing new interpretation.

Where they are combined, the seven moves above still apply in the same order. Check your institution’s research degree regulations and, more usefully, read two or three recent successful theses from your own department — local convention varies more than published advice suggests, and your examiners are likely to share it.

A practical drafting order

Do not write this chapter front to back on the first pass. Draft Move 2 first, finding by finding, because it is the most concrete. Then write Move 4, since the contribution usually becomes clear only once you have interpreted everything. Then Move 3, which depends on both. Write Move 1 last, when you finally know what the chapter argues.

Limitations and future work can be drafted at any point, and are worth writing early — the act of listing what your study cannot show often clarifies what it can.

If your findings are settled and it is the drafting that has stalled, you can build the chapter section by section in Tesify, working from your own results and your own reading of them. The interpretation and the argument remain 100% written by you.

Frequently asked questions

Should the discussion repeat the numbers from my results chapter?

Refer to findings, but do not re-report them in full. Give enough for the reader to follow the argument — the direction of an effect and its approximate magnitude — and cross-reference the results chapter for the detail. Full tables repeated here signal that the chapter is describing rather than interpreting.

Can I introduce new literature in the discussion?

Yes, and it is often necessary. Findings you did not anticipate require literature you did not review at the outset. What you should not do is introduce a body of work so central that its absence from your literature review looks like an oversight.

How do I discuss a null result?

Directly, and with attention to power. Distinguish between evidence of absence and absence of evidence, report the precision of your estimate, and consider methodological explanations alongside theoretical ones. A well-analysed null result is a legitimate doctoral contribution, particularly where the effect was widely assumed.

Is it acceptable to say my findings contradict a well-known study?

Yes, provided you work through the alternative explanations first and you are specific about what differs between the two studies. Examiners respect a candidate who can say clearly why two studies diverge; they are wary of one who dismisses established work without accounting for it.

How many limitations should I include?

Three or four substantive ones, each with its consequence analysed. Quantity is not the measure — a short section that shows you understand how your design shapes your conclusions beats a long list of generic caveats.

Should the discussion chapter be written before or after the introduction?

After. The introduction has to promise what the thesis delivers, and you cannot know precisely what that is until the discussion is drafted. Writing the introduction first is the most common cause of a mismatch between what a thesis claims at the start and what it demonstrates at the end.
August 14, 2026