The methodology chapter is where students lose marks they never needed to lose. Not because their research was poor — often the study itself was perfectly sensible — but because the chapter described what they did without ever explaining why, leaving the marker to guess at the reasoning. Your marker cannot award marks for decisions they cannot see you make. A good methodology chapter does one job: it shows that every choice in your study — the approach, the design, the sample, the analysis — was made deliberately, in light of alternatives, to answer your particular research question. This guide walks through each element in the order your chapter should present it. If you are still structuring the whole document, read it alongside our dissertation structure guide, and if your literature review chapter is still to come, our literature review guide covers that stage.
What the methodology chapter is for
Your research question asks something. The methodology chapter explains how you set out to answer it, and why that way and not another. Two tests run through everything below:
- Alignment. Every element must visibly serve the research question. A question about how many or how strongly points to measurement; a question about how people experience something points to their accounts of it. Markers check this first.
- Justification. For each major choice, name at least one alternative you rejected and say why. “Interviews were used” describes. “Interviews were used rather than a survey because the study sought participants’ reasoning in their own terms, which closed questions cannot capture” justifies — and justification is where the marks are.
One more distinction: methodology is the overall strategy (the approach and design), while methods are the tools (the questionnaire, the interview schedule). Your chapter covers both, strategy first.
Research philosophy, in plain English
Philosophy sections frighten students because textbooks present them as a wall of “-isms”. In practice, at dissertation level you need to locate your study on one simple spectrum and say so honestly.
Positivism treats social reality as something measurable, like the physical world: there are patterns out there, and the researcher’s job is to measure them objectively, ideally with numbers, samples big enough to generalise from, and minimal interference from the researcher’s own views. Surveys, experiments, and statistical studies sit here.
Interpretivism starts from the opposite assumption: what matters about people is the meaning they attach to things, and meanings cannot be counted — they have to be understood, usually by listening. Interviews, focus groups, and case studies sit here, the sample is small because the data is deep, and the researcher’s own interpretation is part of the instrument rather than a contaminant.
Pragmatism refuses the choice: pick whatever combination of methods answers the question. Mixed-methods studies usually claim this ground.
For your chapter, the honest version is short: one or two paragraphs naming your position, defining it in a sentence, and — the step students skip — stating what it ruled in and out for your study. If your dissertation is literature-based with no fieldwork, say so plainly here: your “philosophy” discussion is about how you treated published evidence, and examiners accept that when it is stated rather than disguised.
Choosing a design: the four you will actually use
Survey (questionnaire). Best for breadth: attitudes, frequencies, comparisons across a group. Strengths: efficient, analysable statistically, anonymous by design. Weaknesses: shallow — you learn that people think something, rarely why — and response rates are your eternal enemy. Justify your questions, pilot the instrument, and report the response rate honestly.
Interviews. Best for depth: experiences, reasoning, sensitive topics. Semi-structured is the dissertation norm — a topic guide, the same core questions for everyone, freedom to probe. Strengths: rich data, the ability to follow the unexpected. Weaknesses: time (recruitment, transcription), small samples, and the need to analyse what people said rather than count it.
Case study. Best for one bounded thing studied in depth: one organisation, one ward, one programme, one event — typically combining several data sources (documents, interviews, observation). Strengths: context and completeness. Weaknesses: findings do not generalise in the statistical sense; you defend them as transferable insight instead.
Secondary data / literature-based design. Best — and often only feasible — when primary data is out of reach: systematic-style reviews, analysis of published datasets, policy or document analysis. Strengths: speed, ethics simplicity, scale. Weaknesses: you are limited to what others collected, so your critical appraisal of those sources is your rigour. If your methodology chapter covers a desk-based dissertation, this section must show a proper search and selection method, not merely “articles were found online”.
State the design, then the sentence that earns marks: why this design fits this question better than the main alternative.
Sampling: who, how many, and why them
Describe who was eligible (inclusion and exclusion criteria), how you recruited them, and how many you aimed for and actually got. Then name your sampling logic:
- Probability sampling (random, stratified) — needed if you will claim generalisability; rare in student fieldwork.
- Purposive sampling — participants chosen because they fit the question (the standard for interview studies). Say what made them the right informants.
- Convenience sampling — whoever was accessible. Sometimes unavoidable; if so, say so and carry it into your limitations rather than dressing it up.
Sample size justification differs by design: surveys argue from response needs and basic power considerations; interview studies argue from saturation — you stopped when new interviews stopped adding themes. A small sample defended well beats a big one defended not at all.
Data collection: the replicable account
Write this section so that another researcher could repeat your study from it alone: the instrument (questionnaire items or interview topic guide — full versions in an appendix), where and when data was collected, how long interviews lasted, how they were recorded and transcribed, what the pilot changed, and how non-response or dropouts were handled. Mention practical problems and what you did about them; real studies have them, and markers trust chapters that admit to them.
Data analysis: matching method to data
Quantitative data is usually analysed in SPSS or similar: descriptive statistics first (frequencies, means, distributions), then the tests your question needs — and only tests your data legitimately supports, checking the assumptions (data type, normality, group sizes) rather than running a t-test because it is the one you remember. If your dissertation involves statistical analysis and you want the analysis stage handled or checked, our SPSS data analysis help covers that service.
Qualitative data is most commonly analysed thematically (Braun and Clarke’s six phases is the citation your marker expects): familiarisation, coding, generating themes, reviewing them, defining them, writing up. Whatever approach you use, show your working — an example of coded transcript in an appendix, a table of themes with supporting quotations in the findings — because in qualitative work the visible audit trail is your evidence of rigour.
Literature-based studies analyse by synthesis: your appraisal framework (a recognised critical appraisal tool), your extraction method, and how themes across sources were built.
Whichever applies, the rule is the same: name the approach, cite its source, describe your actual steps, and connect the output to your findings chapter.
Ethics
Every methodology chapter needs an ethics section, and “no ethical issues arose” is never an adequate one. Cover: approval (which committee approved the study, and when — before any data collection); informed consent (participant information sheet, right to withdraw, how consent was recorded); anonymity and confidentiality (pseudonyms or participant codes, what happens to anything identifiable); data storage (password-protected, retained and destroyed per university rules and GDPR); and any sensitive-topic handling (distress protocol for interviews, signposting support). Literature-only studies state briefly that no human participants were involved and approval was therefore not required — briefly, not silently.
Reliability, validity — and trustworthiness
Quantitative chapters defend reliability (would the instrument give consistent results? pilot data, Cronbach’s alpha where relevant) and validity (does it measure what it claims? face/content validity, using established instruments where they exist).
Qualitative chapters should not borrow those terms uncomfortably; the standard vocabulary is trustworthiness: credibility (prolonged engagement, member checking, triangulation), transferability (thick description so readers can judge applicability), dependability and confirmability (the audit trail, reflexivity — your honest account of how your own position may have shaped the study). A short reflexivity paragraph is one of the cheapest credibility wins available in any qualitative dissertation.
Limitations
End the chapter (or place it here if your university prefers, rather than in the conclusion) with the study’s real limitations: sample size and who is missing from it, single-site scope, self-report bias, cross-sectional timing, your own novice-researcher status where it genuinely mattered. For each, one sentence on what it means for the findings. Markers do not punish honest limitations; they punish discovering limitations you did not declare.
A structure you can copy
- Introduction — restate the aim and research question; roadmap of the chapter
- Research philosophy and approach
- Research design — with the rejected alternative justified away
- Setting and participants/sample — criteria, recruitment, numbers
- Data collection — instruments, procedure, pilot
- Data analysis — approach and actual steps
- Ethics — approval, consent, confidentiality, storage
- Reliability/validity or trustworthiness
- Limitations
- Summary — how the methodology answers the question set in section 1
Common mistakes
Writing it afterwards from memory. Decisions blur; keep a research diary from day one and the chapter largely writes itself.
Describing without justifying. The single most common reason methodology chapters underperform — every major choice needs its “why”.
Philosophy pasted from a textbook. Three pages defining ontology that never touches your study scores less than three honest paragraphs that do.
Method–question mismatch. Interviews bolted onto a “how many” question, or a 20-person survey claiming to measure a population. Alignment is checked first.
Ethics as an afterthought — or worse, data collected before approval. Approvals come first, always; unapproved data is normally unusable.
Hiding the mess. Non-response, a failed recruitment route, a question everyone misunderstood: declared and handled, these read as competence. Concealed, they read as carelessness when the marker spots them — and markers spot them.
Frequently asked questions
How long should a methodology chapter be?
As a rough guide, 10–15% of the dissertation’s total word count — around 1,000–1,500 words in a 10,000-word dissertation. Proportion matters more than the number: a complex mixed-methods study needs more space than a straightforward survey.
What tense should it be written in?
Past tense for what you did (“participants were recruited…”), present tense for established methodological claims (“thematic analysis involves…”). Your university’s guide may specify a voice preference — third person is the traditional default.
My dissertation is literature-based. Do I still need this chapter?
Yes — a shorter one. Your philosophy, your search and selection method, your appraisal framework, and your synthesis approach take the place of fieldwork sections. Desk-based is a design, not an absence of one.
Method first, marks follow
A methodology chapter is not bureaucracy between your question and your findings — it is the evidence that your findings deserve to be believed. Write it as a chain of justified decisions, keep every link aligned to the research question, and the chapter does its job. If you want experienced eyes on yours before submission, send it to us on WhatsApp or email team@unitedacademichelp.uk — dissertation support from £20 per 1,000 words, with no obligation.