How to Write a Dissertation Methodology Chapter That Earns Marks

I need to write my dissertation methodology. What actually goes in it? Do I explain my choices or just list what I did? How much detail is enough, and where do ethics fit?

Those four questions come up at almost every methodology meeting, and they have clear answers. A dissertation methodology is the chapter where you show how you answered your research question and why that route was sound. Markers are not checking whether you picked the “right” method. They are checking whether your choices fit your question and whether someone else could follow your steps.

This covers dissertations that collect new data (surveys, interviews, experiments) and dissertations built on existing data or literature. The order of sections below suits most social science, business, health and education projects. Science and humanities students should adjust it, and a later section shows how.

What a dissertation methodology chapter must do

The chapter has two jobs. It describes what you did, and it defends why you did it that way. A list of steps with no reasons reads as description. A chapter of reasons with no steps reads as opinion. Strong chapters do both in every section.

Methodology and methods are related but not the same. Methodology is your overall logic: the philosophy, approach and design behind the study. Methods are the specific tools you used, such as a questionnaire, a semi-structured interview or a regression test. Your university library guides usually draw this line, and the University of Southampton library methodology guide is a good example of how a UK library frames it. Most chapters need both, so keep the reasoning and the tools together.

Your chapter should also match your own discipline’s conventions. Some departments want a short methods section inside a larger chapter, while others expect a full methodology chapter of several thousand words. Check your module handbook and marking criteria first, because those override any general advice.

Before you write: five decisions to settle

A methodology chapter is easy to write once the decisions behind it are made. If a section feels impossible, one of these is usually unsettled.

• What exactly is my research question, and does it ask “what”, “why”, “how much” or “what is the relationship”?
• Do I need numbers, words, or both?
• Will I collect new (primary) data, use existing (secondary) data, or review literature?
• Who or what will I study, and how can I reach them within my time and ethics limits?
• What could go wrong, and how will I show the study is trustworthy?

If the first answer is shaky, fix it before anything else. The guide on how to write a research question helps you tighten it, and the research question generator gives you wording to test against your idea.

How to structure your dissertation methodology, section by section

Use these sections as a default order and drop any that your department does not want. Each one answers a question your marker will silently ask.

Restate the aim and link it to your approach

Open with one or two sentences that remind the reader of your research question and the overall approach. This anchors everything that follows. Do not repeat your whole introduction. If you are still shaping that part, see the guide on how to write a dissertation introduction.

Explain your research philosophy and approach

Research philosophy covers your assumptions about knowledge. Two terms matter most. Ontology asks what reality is, and epistemology asks how we can know about it. Positivism treats knowledge as measurable and objective, interpretivism treats it as shaped by people’s meanings, and pragmatism lets the question decide the method.

Then state your approach to theory. A deductive approach tests an existing theory or hypothesis. An inductive approach builds themes or theory from the data. An abductive approach moves between the two. Many business and management courses in the UK use the “research onion” from Saunders and colleagues to lay these layers out in order, so check whether your supervisor expects it.

Keep this to what your project needs. A short, accurate paragraph beats a philosophy lecture. For example, a student studying how nurses experience night shifts would usually explain an interpretivist stance and an inductive approach, because the aim is to understand meaning rather than test a hypothesis.

Choose and justify your research design

Research design is the plan that connects your question to your data. Name it, then say why it fits. Common options include a cross-sectional survey, an experiment, a case study, a longitudinal study and a phenomenological study.

Your design usually follows from the type of data you need.

• Quantitative research collects numerical data to measure, compare or test relationships.
• Qualitative research collects words, images or observations to explore meaning and experience.
• Mixed methods combines both. In a sequential design one strand informs the next, and in a concurrent design both run together.

Say plainly why you did not choose the alternatives. “A survey could not capture why staff left, so interviews were added” tells the marker you weighed your options.

Describe your data sources and sampling

Say whether your data is primary (collected by you) or secondary (collected by someone else). Then explain who or what you studied, called your population, and how you selected a sample from it.

Probability sampling gives every member of the population a known chance of selection, which supports generalising to the population. Non-probability sampling, such as purposive, snowball or convenience sampling, suits studies that need specific people or hard-to-reach groups, but it limits generalisation. If you use a non-probability sample, say so openly and discuss the effect in your limitations.

Give your sample size and the reason for it. For quantitative work, explain how you reached the number, and the dissertation sample size calculator can help you check your reasoning. For qualitative work, avoid saying a number is “enough” without a reason. Guest, Bunce and Johnson (2006) found that data saturation appeared by around twelve interviews in a fairly uniform group, but that finding is not a universal rule, and your own design and population decide what is adequate. Say how you judged that your data was sufficient.

Set out your data collection tools and procedure

Describe each tool and how you used it. For a questionnaire, state how questions were developed or adapted, the response format, and whether you piloted it. For interviews, state whether they were structured, semi-structured or unstructured, how long they ran, how they were recorded and how they were transcribed. For secondary data, name the dataset, the owner, the period covered and the variables you used.

Write the procedure so a reader could repeat it: what you did first, what came next, and over what period. Put the full questionnaire or interview guide in the appendices and refer to it from the text rather than pasting it into the chapter.

Explain how you analysed the data

Name the technique and justify it. Match the method to the data and the question.

• Thematic analysis suits patterns of meaning in interview or open-text data. Braun and Clarke describe six phases, from familiarisation to writing up, and their later reflexive approach stresses that themes are built by the researcher rather than “found”.
• Content analysis suits documents, media and text where you code and count categories.
• Descriptive statistics, correlation, regression and group comparison tests suit numerical data.
• Software such as SPSS or NVivo can be named, but the software is a tool, not a method. Say what you did with it.

If you are unsure which test fits, the statistical test selector narrows the options. State any assumptions you checked before running a test.

Show validity, reliability and trustworthiness

Quantitative studies are judged on validity (does it measure what it claims to measure?) and reliability (would it give consistent results?). If you use a scale of several items, a Cronbach’s alpha is a common reliability check. Tavakol and Dennick (2011) note that values from about 0.70 to 0.95 are generally considered acceptable, and that very high values can mean items repeat each other.

Qualitative studies use different language. Lincoln and Guba proposed credibility, transferability, dependability and confirmability. Ways to build these include keeping an audit trail, checking interpretations with participants, and writing a short reflexive note on how your own background may shape your reading of the data.

Cover research ethics clearly

Ethics belongs inside the methodology, not only in an appendix. Cover informed consent, the right to withdraw, confidentiality or anonymity, safe storage of data, and any risk to participants or to you. Say that you received approval from your university ethics committee, and include the approval reference if your department asks for it. Do not start collecting data before approval is confirmed.

Several bodies publish ethics guidance that UK supervisors often point students towards, including the British Educational Research Association (BERA) ethical guidelines for educational research, fifth edition, 2024, and the British Psychological Society (BPS) Code of Human Research Ethics. Use the one your supervisor names.

Health and social care students should also check NHS rules. The Health Research Authority (HRA) says that from 1 September 2021 it changed how student research is handled, so that undergraduate student projects generally do not need NHS research ethics committee or HRA review, some Master’s projects can follow a student research toolkit route, and doctoral research is unchanged. Read the current HRA student research guidance and then confirm with your supervisor, because your university may still require its own review.

Data protection also applies when you hold personal data. UK data protection law has been updated by the Data (Use and Access) Act 2025, and most of its research-related changes were brought into force in February 2026. The ICO’s guidance on the research provisions is under review at the time of writing. Treat the ICO research provisions guidance as the place to check, and follow your university’s data protection policy. Position correct as of October 2026; check the HRA and ICO sources above for the current status.

State your limitations honestly

Every study has limits, and a marker trusts a student who names them. Link each limitation to your design: a small non-random sample, a short time window, self-reported answers or one case. Then say what you did to reduce the effect, and avoid apologising. Limitations belong in your methodology, and you can return to them in the conclusion.

A short worked example

A Master’s student wants to know why mid-career staff leave a regional hospital trust. The question asks “why”, so the student chooses an interpretivist stance, an inductive approach and a qualitative case study design.

Sampling is purposive: staff who left within the past two years and agreed to take part. Data comes from semi-structured interviews, recorded with consent and transcribed. Analysis uses reflexive thematic analysis. Trustworthiness is built through an audit trail and a reflexive log. Ethics approval comes from the university committee, and because the participants are NHS staff rather than patients, the student checks with the supervisor whether any NHS review applies.

Each choice has a reason that traces back to the question. That chain from question to design to sampling to analysis is what a methodology chapter should show. This is an illustration, not a template to copy, because your own question must lead your choices.

Common methodology mistakes and how to fix them

Mistake	Why it costs marks	Fix
Describing methods with no reasons	Reads as a list, not an argument	Add “because” to every choice and link it to the question
Choosing a method first, then a question	The design does not fit	Rewrite the question or change the method
Writing in future tense	The work is finished	Use past tense, unless the chapter is for a proposal
Mixing “I” and “the researcher”	Reads as inconsistent	Pick one style and ask your supervisor which they prefer
Copying a textbook’s definitions at length	No sign of your own thinking	Define a term briefly and apply it to your study
Missing ethics or approval details	A basic compliance gap	State approval, consent, anonymity and data storage
Putting results or findings in this chapter	Belongs in the next chapter	Keep this chapter to method and save results for later

Methodology for dissertations without new data

If you use only existing material, you still need a methodology. State how you chose your sources, what search terms and databases you used, what you included and excluded, and how you judged quality. A systematic or structured literature review needs a clear search and selection process, and a secondary data study needs the dataset details and analysis steps. A short guide to how to write a literature review covers the searching and synthesis side.

Humanities and creative subjects often work through close reading, archival work, practice-based enquiry or a critical framework instead of sampling. The same principle applies: name your approach, justify it, and show how you handled your sources. Laboratory and applied science projects usually add equipment, materials, protocols and calibration, where the key test is that another researcher could repeat the work.

Tense, length and presentation

Write the chapter in the past tense because the work is done. Choose first person or third person and stay with it. Keep paragraphs short, use the headings your department expects, and refer to appendices for questionnaires, consent forms and interview guides. Length varies by level and department, so check the word limit and marking criteria for your course rather than relying on a typical figure.

Plan time for this chapter early. Ethics approval and participant recruitment can take longer than writing, and the dissertation timeline planner helps you place those steps before your deadlines. If you plan to build on a proposal, the guide to how to write a dissertation proposal shows how its method section can grow into this chapter.

Final checklist before you submit

• Does every method link back to the research question?
• Have I justified each choice and said why I rejected the main alternative?
• Could another researcher repeat my steps from this chapter and the appendices?
• Are sampling, sample size and the data collection procedure clear?
• Have I named the analysis method and the steps taken?
• Do I explain validity, reliability or trustworthiness in the right language for my design?
• Are ethics, consent, data storage and approval covered?
• Have I stated limitations and how I reduced their effect?
• Is it in the past tense, with a consistent voice, and no results in it?

Questions students ask about the methodology chapter

Can I use “I” in my dissertation methodology?

It depends on your discipline and supervisor. Many social science and humanities courses accept first person, especially for reflexive qualitative work, while some science departments prefer third person. Ask your supervisor, check your handbook, and stay consistent.

Should the methodology chapter include a literature review?

A brief reference to the literature that supports your method is fine, such as the source of a scale or the authors behind an analysis approach. The main review of your topic belongs in your literature review chapter.

What if my method did not work as planned?

Report it honestly. Explain what changed, why, and how it affected your data. Markers value clear reasoning about problems more than a tidy account that hides them. Speak to your supervisor early when a plan fails.

Is it acceptable to get help with the chapter?

Academic support such as guidance on structure, feedback and editing can help you understand what a strong chapter needs. If you want tailored help, dissertation methodology help is available, and any support should be used for learning and according to your institution’s policy.