Dissertation methodology mentoring.

Almost nobody fails a dissertation because they picked the wrong method. People lose marks because they cannot explain why they picked it — why interviews rather than a survey, why thirty participants rather than three hundred, why thematic analysis rather than framework analysis, and what each of those choices cost them. A methodology chapter is an argument, not a description of what you did.

That argument is what we work on. A mentor reads what you have, asks the questions your marker or your examiner will ask, and keeps asking until you can answer them in your own words. Then you write the chapter. We are a second reader and a hard questioner, which is the thing most students are short of and no amount of reading replaces.

What methodology mentoring is, and what we will not do

Your methodology is the one part of a dissertation that has to survive being questioned out loud. In a taught Master's that questioning happens on the page, through a marker's comments against the assessment criteria. In a doctorate it happens in a room, with two examiners and your thesis open in front of them. Either way, the test is the same: can you account for your decisions?

Work bought from somebody else cannot pass that test, which is the practical reason we coach rather than write, as well as the principled one. Our academic integrity policy sets out the boundary in detail. In short:

One test, if you are unsure whether what you want is on the right side of that line: could you describe exactly what help you had to your supervisor, and have the conversation be unremarkable? If yes, it is mentoring. If you would need to leave something out, it is not.

The question your methodology chapter actually has to answer

Not “what did you do”. The question is: why is this the right way to answer your research question, and what are you therefore entitled to claim at the end?

That framing matters because it makes the chapter answerable. Every design decision gets the same three-part treatment:

1. The choice. What you did, in one sentence, precisely enough to be repeated.
2. The reason. What that choice makes possible that the obvious alternative does not.
3. The cost. What you gave up, and where in the thesis you account for it.

The third part is the one students skip, and it is the one that reads as mature. A chapter that names its own limits is far stronger than one that pretends there were none — and at doctoral level it is more or less required, because the UK doctoral standard is an original contribution made through independent research, and independence shows up as judgement about trade-offs rather than as a flawless design.

Where methodology chapters lose marks

Six patterns recur across disciplines and levels. None of them is about picking a wrong method.

1. The method was inherited, not chosen. The design came from the last paper you read, or from what your supervisor's other students did, and the chapter reads as a defence constructed afterwards. Markers and examiners spot this immediately, usually by asking what else you considered.
2. Philosophy is decorative. Two pages on positivism and interpretivism that never touch the actual design. Either your epistemological position does work in the argument — it explains why this evidence counts as evidence — or it should be one paragraph, not a section.
3. The question and the method do not match. A question asking “how do people experience X” answered with a Likert-scale survey; a question asking “does X cause Y” answered with six interviews. Fixable in week two, expensive in month six.
4. Sampling is reported but not justified. “A convenience sample of 42 participants was used.” Why 42? What does 42 let you say, and what does it stop you saying?
5. The vocabulary belongs to the other tradition. Claiming “validity and reliability” for a small qualitative study, or “trustworthiness” for a regression model. See validity, reliability, trustworthiness below.
6. What went wrong is hidden. Recruitment fell short, the instrument did not work, access was withdrawn — and the chapter describes the plan rather than the study. Reporting the deviation and handling it is worth more than a tidy fiction, and in a viva it is the difference between a short conversation and a long one.

Choosing a research design: the trade-off, not the label

Students ask which design is best. There is no answer to that, and the question itself is a symptom — it means the research question has not been settled yet. What there is, is a set of trade-offs you can reason about.

Quantitative designs

Appropriate when your question is about how much, how many, how often, or whether a relationship holds across a population. They buy you breadth and the ability to generalise within stated limits; they cost you depth and context. The questions to be ready for: where did your instrument come from and has it been validated; how did you arrive at your sample size; what are you controlling for and what are you not; and are you claiming correlation or causation, because a cross-sectional survey cannot give you the second one.

Qualitative designs

Appropriate when your question is about meaning, process, experience or mechanism — how and why rather than how many. They buy you depth and the ability to see something the literature has not framed yet; they cost you reach. The questions to be ready for: why this approach rather than another with the same data (a thematic analysis, a framework analysis, an IPA study and a grounded-theory study are not interchangeable and the choice has to follow from the question); how you moved from transcript to code to theme; and what your own position did to the analysis.

Mixed methods, and how to justify it properly

Mixed methods is the design most often chosen for the wrong reason — because it sounds more thorough, or because the student could not decide. In a dissertation with a fixed word count it is also the design most likely to produce two thin studies instead of one good one. It is worth doing when, and only when, you can say what the second strand does that the first cannot.

There are only a few honest answers to that, and naming yours is the justification:

• Sequential explanatory. The quantitative strand finds a pattern; the qualitative strand explains why it is there. You must be able to say what you would have asked if the pattern had come out differently.
• Sequential exploratory. The qualitative strand tells you what to measure, because the existing instruments do not fit your context. You must show the instrument was genuinely built from your findings.
• Convergent. Two strands address the same question independently so that agreement strengthens the claim. You must say in advance what you will do if they disagree — and disagreement is a finding, not a failure.
• Embedded. One strand is subordinate and serves a specific purpose, such as a process evaluation inside a trial. You must be clear which is which, so the reader is not expecting symmetry.

Then the practical question nobody asks until it is too late: where do the strands actually meet? If the integration happens only in the discussion, in a paragraph beginning “both strands suggested…”, you have written two studies and stapled them. Plan the point of integration at design stage and say in the methodology where it will be.

Sampling, and the claim it buys you

Sampling is where most methodology sections quietly overreach. The rule is simple and it is the one thing to take from this section: your sample sets the ceiling on your conclusion. Everything in the discussion has to sit under that ceiling.

Three things a marker or examiner will check:

1. Is the sample adequate to the claim? Not “is it big”. A purposive sample of eight can support a strong claim about a mechanism and no claim at all about prevalence. A convenience sample of 400 students can support a claim about students at one institution and not about a sector.
2. Is recruitment reported transparently? How you reached people, who said no, who dropped out, and what that does to the composition. Non-response is information, not embarrassment.
3. Are the limits acknowledged in the right place? In the methodology where you set the boundary, and again in the discussion where you keep to it. A limitations paragraph that contradicts the conclusion two pages later is worse than no limitations paragraph.

On sample size: at Master's level there is rarely a published rule you can cite, and inventing one is worse than reasoning out loud. For quantitative work, show the calculation or say frankly that the sample was constrained by access and what that means for power. For qualitative work, justify by the logic of the design — information richness, case-selection strategy, or reaching the point where new data stopped changing the analysis — and only claim saturation if you can describe how you noticed it.

Validity, reliability, trustworthiness: use the right vocabulary

Two research traditions, two sets of words, and using the wrong set signals a shallow reading of your own methods literature faster than almost anything else. This is cheap to get right.

Positionality belongs in this section too, and it is now expected in social sciences, education, health and increasingly elsewhere. A positionality statement is not a biography and it is not an apology. It says what about your position gave you access or shaped interpretation, and what you did about it. Two specific sentences beat a page of disclosure.

Ethics and data management: the part that delays people

Ethics approval is the single most common cause of a dissertation timetable slipping, because it is the one part of the process you do not control. The committee meets when it meets. Plan backwards from its deadlines, not forwards from yours.

• Find your committee's dates before you design the study. Departmental, school and university-level committees have different thresholds and different calendars, and work involving the NHS, children, prisoners or other groups needing additional protection goes through a separate and longer route.
• Write the application as part of the design, not after it. The questions it asks — what are you collecting, from whom, on what basis, how will you store it, when will you destroy it — are methodology questions. Answering them early improves the chapter.
• A data management plan is not paperwork. Where the data lives, who can see it, how it is pseudonymised, what happens to it at the end. Examiners ask about this, and at doctoral level funders frequently require it.
• Consent has to match what you will actually do. If you might quote participants, say so. If you intend to deposit the data, say so at consent stage, because you cannot add that afterwards.

Your institution's own research ethics policy and your funder's conditions govern all of this, and they override any general guidance including ours. The sector-level framework is worth reading once so you understand what your committee is working to — see sources.

By level: undergraduate, taught Master's, MRes, doctoral

The methodology chapter does a different job at each level, and expectations differ more than most students realise.

Undergraduate dissertation

Usually a smaller study with a tighter word count and often secondary data or a single accessible sample. The chapter is assessed mainly on whether you understood the choice you made. Depth of methods literature matters less than coherence between question, design and conclusion.

Taught Master's dissertation (MA, MSc, MBA, LLM)

The volume case. Typically 12,000–20,000 words, a single empirical study or systematic review, and three to five months of real working time inside a one-year programme. The marking criteria usually reward explicit justification and explicit limitations, which is exactly where mentoring earns its keep. Deadlines cluster in August and September, so methodology decisions want to be settled by late spring.

MRes and Master's by research

Longer, more methodologically ambitious, and sometimes examined by an oral as well as a thesis — it varies by institution, so check your programme handbook rather than assuming. Expect to defend design choices at something closer to doctoral depth on a Master's timetable.

Doctoral thesis (PhD, professional doctorates)

Here the methodology is not a chapter that reports a decision; it is part of the contribution. You will be asked why this design and not the alternative, what happened when it met reality, and what you would do differently — and the answers are examined out loud. Many institutions also test the design mid-programme, at an upgrade, transfer or confirmation panel, before the thesis exists. If that is where you are, start with upgrade and transfer viva preparation, which is about defending a plan rather than a finding.

By discipline

What counts as a well-justified method is partly a disciplinary convention. A reproducibility argument that satisfies an engineering examiner will not satisfy a sociologist asking about positionality, and an OSCOLA-referenced doctrinal legal analysis is not a “qualitative study” however much it looks like one from outside. We match you to a mentor inside your own discipline for exactly that reason.

The nine specialist areas we mentor in are below. Dedicated guidance pages for each discipline are in preparation and are not published yet — the items below are deliberately plain text, not links, so that nothing here promises a page that does not exist. To talk to a mentor in one of these areas now, use the quote form and name your field.

How methodology mentoring works

1. You send what you have. The research question, the module or programme handbook with its assessment criteria, and whatever draft exists — even if it is three paragraphs and a worry.
2. We match you to a mentor in your discipline. Not a generalist. Someone who knows what your field accepts as evidence.
3. First session: the question. Before design, we establish whether your question is answerable with the time, word count and access you actually have. This session changes the project more often than any other.
4. Design sessions. Options, trade-offs, and you arguing for one. You leave with notes you wrote, not a document we wrote.
5. Written feedback on your draft chapter. Commentary on the reasoning: which justifications hold, which are assertions, where the claim exceeds the sample, where the vocabulary slips between traditions. Marginal comments and a summary — never tracked rewrites of your sentences.
6. A rehearsal. Your mentor asks the questions a marker or examiner would, and you answer them out loud. For doctoral candidates this extends into full mock viva preparation.

Sessions are online and scheduled around your deadlines. Mentors work Monday to Saturday, 9am to 10pm GMT. Everything you send stays confidential, and we keep a record of your drafts so the development of your own work is documented if your institution ever asks.

Defending your methodology when you are questioned on it

Methodology is where most vivas spend most of their time, and the questioning follows a predictable shape — why this design, justify your sampling, what went wrong in the field, how did you establish rigour, what is your positionality. Each one is a check on a specific thing, and knowing which check is being run changes how you answer.

We have written that out question by question in our guide to how examiners question a methodology chapter, which shows what each question is testing, what a weak answer sounds like and the follow-up it invites. If your viva is already in the diary, the practical next step is PhD viva preparation coaching.

Pricing

Three tiers, priced per word of the work being mentored, in pounds sterling. The same prices apply to every student wherever they are reading this; there is no regional pricing and no currency switching.

The quote form gives you a figure before you speak to anybody, and nothing is submitted until you choose to send it.

Your mentor

Questions students ask before booking

Related guidance

Sources

The sector-level documents this page works from. Every link was checked on 8 October 2026.