A practical, step-by-step guide for postgraduate students — covering research philosophy, the Saunders Research Onion, qualitative vs. quantitative methods, sampling, analysis, and the mistakes that lose marks.
Your methodology chapter is where examiners decide whether to trust your results. You can have a brilliant topic and elegant writing, but if your methods don’t logically connect to your research questions, the whole dissertation wobbles. This guide walks you through how to choose a methodology that holds up — using the frameworks UK universities actually expect you to reference, with worked examples and a decision checklist you can apply to your own project.
By the end, you’ll be able to justify every methodological choice you make, which is exactly what the marking criteria reward. If you get stuck, our master’s dissertation help service can review your chapter, but this guide gives you everything you need to do it yourself.
Why the Methodology Chapter Carries So Much Weight
In most UK master’s mark schemes, methodology accounts for a large share of the grade — often more than the literature review — because it demonstrates research competence rather than just reading. Examiners aren’t only asking “what did you do?” They’re asking “why was this the right way to do it, and can I trust the findings that follow?”
A well-argued methodology chapter does five things:
- Answers your research questions through a defensible plan, not assumptions
- Justifies each choice with academic reasoning and references
- Demonstrates reliability and validity (or, in qualitative work, trustworthiness)
- Acknowledges limitations honestly, which signals maturity rather than weakness
- Creates a logical thread from philosophy → design → data → analysis → ethics
The single most common reason students lose marks here is description without justification. Saying “I used semi-structured interviews” earns little; explaining why interviews suited your interpretivist aims, and why semi-structured ones in particular, is what moves you up a grade band.
The Saunders Research Onion: The Framework UK Examiners Expect
If you study at a UK university, you have almost certainly been pointed toward the Saunders, Lewis and Thornhill “Research Onion” (from Research Methods for Business Students). It’s the framework most supervisors expect you to map your methodology against, so it’s worth structuring your chapter around its layers — working from the outside in:
| Layer | What you decide | Common options |
|---|---|---|
| 1. Philosophy | Your beliefs about knowledge and reality | Positivism, Interpretivism, Pragmatism, Realism |
| 2. Approach | How theory and data relate | Deductive, Inductive, Abductive |
| 3. Strategy | Your overall plan for answering the question | Survey, Case study, Experiment, Grounded theory, Ethnography, Action research |
| 4. Choices | How many method types you combine | Mono-method, Mixed methods, Multi-method |
| 5. Time horizon | The timeframe of your study | Cross-sectional (snapshot) or Longitudinal (over time) |
| 6. Techniques | How you collect and analyse data | Sampling, data collection, data analysis |
Walking your reader through these layers in order gives your chapter a logical spine and shows examiners you understand methodology as a connected system, not a checklist.
Layer 1 — Choosing Your Research Philosophy as the best research methodology
Your philosophy shapes everything downstream, so settle it first. It reflects your assumptions about what counts as valid knowledge (epistemology) and the nature of reality (ontology).
Positivism
Positivism treats reality as objective and measurable. You test theories using numerical data and statistical analysis, aiming for findings that are generalisable and replicable. It’s common in finance, economics, operations, and quantitative marketing research.
Example: measuring whether social-media ad spend predicts sales across 200 firms using regression analysis.
Interpretivism
Interpretivism focuses on understanding the meanings, experiences, and perceptions people attach to a situation. Reality is seen as socially constructed, so the goal is depth of understanding rather than generalisable laws. It dominates education, sociology, HR, and much of qualitative business research.
Example: exploring how first-year students experience the transition to online learning, through in-depth interviews.
Pragmatism
Pragmatism sidesteps the philosophical either/or and asks “what best answers my question?” It comfortably combines qualitative and quantitative work and underpins most mixed-methods dissertations. If your research questions have both a “how many/how much” and a “why/how” element, pragmatism is often the honest fit.
Realism (Critical Realism)
Realism sits between positivism and interpretivism: it accepts an objective reality exists but argues our understanding of it is shaped by social context. It’s increasingly popular in management and policy research where structures and individual perceptions both matter.
Tip: Don’t pick a philosophy because it sounds impressive. Pick the one your research questions already imply, then justify it. Examiners spot retro-fitted philosophy immediately.
Layer 2 — Your Research Approach
| Approach | Starts with | Best for | Typical pairing |
|---|---|---|---|
| Deductive | An existing theory you test | Hypothesis-driven, quantitative studies | Positivism, surveys, statistics |
| Inductive | Observations you build theory from | Exploring under-researched areas | Interpretivism, interviews, thematic analysis |
| Abductive | Moving back and forth between data and theory | Real-world business problems | Pragmatism, mixed methods |
Most taught-master’s dissertations in business and management lean deductive; most in education and the social sciences lean inductive. Either is fine — what matters is that the approach matches your philosophy and questions.
Layer 3 — Your Research Strategy (and Design Type)
Strategy is your overall plan of attack. Beyond the familiar survey and experiment, UK students frequently use:
- Case study — an in-depth investigation of one organisation, group, or event in its real context. Ideal when you want rich detail rather than breadth.
- Grounded theory — building theory directly from data through iterative coding. Demanding but powerful for genuinely new areas.
- Ethnography — immersing yourself in a setting to understand culture and behaviour. Time-intensive, so rarely full-scale at master’s level.
- Action research — researching and changing a practice in cycles, common in education and healthcare.
It also helps to classify your design by purpose:
- Exploratory — for topics with little prior research; you’re scoping and finding patterns.
- Descriptive — for mapping characteristics, trends, or relationships in a population.
- Explanatory — for testing cause and effect, asking why something happens.
Qualitative vs. Quantitative vs. Mixed Methods
This is the choice most students agonise over, so here it is side by side:
| Qualitative | Quantitative | Mixed methods | |
|---|---|---|---|
| Goal | Understand meaning and experience | Measure and test relationships | Both depth and measurement |
| Data | Words, observations, transcripts | Numbers, scores, statistics | Both |
| Typical methods | Interviews, focus groups, observation | Surveys, experiments, secondary datasets | Survey + interviews |
| Sample | Smaller, purposive | Larger, often random | Combination |
| Analysis | Thematic, content, narrative | Descriptive & inferential statistics | Both, often via triangulation |
| Strengths | Rich, contextual, flexible | Generalisable, objective, comparable | Balanced, robust, triangulated |
| Limitations | Researcher bias, harder to generalise | Misses nuance and context | More time and planning required |
Qualitative Methods in Brief
Interviews give detailed personal insight and come as structured, semi-structured, or unstructured. Semi-structured interviews are the workhorse of master’s research because they keep a consistent direction while allowing you to follow up on what participants actually say. Focus groups surface collective views and the way opinions form through interaction. Observation captures behaviour in its real setting, either as a participant or non-participant.
Quantitative Methods in Brief
Surveys/questionnaires efficiently gather standardised data from large samples — the most common quantitative tool at master’s level. Experiments isolate cause and effect by controlling variables. Secondary numerical data (government statistics, the ONS, financial databases, company reports) lets you run rigorous analysis without collecting primary data, which is a smart, realistic route when time is tight.
Why Mixed Methods Is Increasingly Popular
Combining a survey (for the “how widespread?”) with interviews (for the “why?”) lets you triangulate — corroborating findings across data types strengthens your conclusions. The trade-off is that you’re effectively running two mini-studies, so only choose it if your timeline genuinely allows.
Sampling: Choosing Who or What You Study
| Type | How it works | Examples | Best for |
|---|---|---|---|
| Probability | Everyone has a known, equal chance of selection | Simple random, stratified, cluster | Quantitative work needing generalisability |
| Non-probability | Selection by judgement, access, or relevance | Purposive, convenience, snowball | Qualitative and exploratory work |
In qualitative research, purposive sampling (deliberately choosing participants who can give the richest, most relevant insight) is standard — not “purpose sampling,” a common typo worth avoiding in your own writing.
On sample size, resist the urge to chase big numbers. Quality of fit beats quantity: 10–15 well-chosen interviewees can produce a stronger qualitative study than 50 poorly matched ones, while quantitative work needs enough responses for your statistical tests to be meaningful (your supervisor or a power calculation can guide the threshold).
Data Analysis: Turning Data Into Findings
Analysing Qualitative Data
The dominant technique is thematic analysis — most often Braun and Clarke’s six-phase approach: familiarising yourself with the data, generating initial codes, searching for themes, reviewing them, defining them, and writing up. You can also use content analysis (counting and categorising) or narrative analysis (examining stories and sequence). Software such as NVivo helps you organise and code larger datasets.
Analysing Quantitative Data
Your toolkit scales with your questions:
- Descriptive statistics — means, frequencies, distributions to summarise the data
- Correlation — whether and how strongly two variables move together
- Regression — the effect of one or more variables on an outcome
- Significance tests — t-tests, ANOVA, chi-square to test hypotheses
SPSS and Excel are the usual tools at this level; R is a strong free alternative if you’re comfortable with code. Whatever you use, name the specific tests and justify them — this is where rigour shows.
Reliability, Validity, and Trustworthiness
For quantitative work, examiners look for reliability (would repeating the study give consistent results?) and validity (are you measuring what you claim to measure?).
For qualitative work, the equivalent is trustworthiness, usually discussed through Lincoln and Guba’s four criteria: credibility, transferability, dependability, and confirmability. Mentioning the right framework for your method type is a quick, easy way to demonstrate that you understand the standards your work will be judged by.
Research Ethics
UK universities take ethics seriously, and most studies involving people require ethical approval before data collection begins. Build these in from the start:
- Informed consent — participants understand the study and agree freely
- Voluntary participation — and the right to withdraw at any time
- Confidentiality and anonymity — protecting identities in your write-up
- Data protection — secure storage and handling in line with UK GDPR
- Avoiding harm — physical, psychological, or reputational
Reference your university’s ethics policy and note when approval was granted; it’s an easy mark and a genuine requirement.
A Worked Example You Can Adapt
Research question: How does remote working affect employee productivity in UK technology firms?
This question has two halves — a measurable “how much” and an experiential “how and why” — so a single method would only tell half the story. A defensible methodology:
| Decision | Choice | Why |
|---|---|---|
| Philosophy | Pragmatism | The question needs both numbers and meaning |
| Approach | Abductive | Test existing productivity theory and stay open to new themes |
| Strategy | Mixed methods (survey + interviews) | Breadth from the survey, depth from interviews |
| Time horizon | Cross-sectional | A snapshot fits a one-year dissertation timeline |
| Sampling | Stratified for the survey; purposive for interviews | Representative numbers, then rich follow-up cases |
| Analysis | Regression for survey data; thematic analysis for interviews | Matches each data type |
| Ethics | Informed consent, anonymity, secure storage, university approval | People-based research |
Notice that every row has a justification. That column is what earns the marks — copy this habit into your own chapter.
The Mistakes That Lose Marks (and How to Avoid Them)
- Describing without justifying. For every method, write the why, not just the what.
- Method–question mismatch. If you want to understand experiences, don’t run a survey; if you want to measure prevalence, don’t rely on three interviews.
- Weak or unexplained sampling. State your sampling type, your inclusion criteria, and your reasoning.
- Ignoring limitations. Every study has them; naming them shows insight, not weakness.
- No academic references. Anchor your choices in methodology texts (Saunders, Bryman, Creswell, Braun & Clarke). An unreferenced methodology chapter reads as opinion.
- Retro-fitting philosophy. Choose it from your questions, not for show.
How to Choose Your Methodology: A Quick Checklist
Work through these in order and your methodology will largely write itself:
- What exactly are my research questions? They drive everything below.
- Do I need numbers, meaning, or both? → quantitative, qualitative, or mixed.
- Which philosophy do my questions imply? Positivism, interpretivism, pragmatism, or realism.
- Deductive or inductive? Testing theory, or building it?
- What’s the best strategy? Survey, case study, grounded theory, etc.
- Who/what is my sample, and how will I select it?
- How will I analyse the data, and with which tools/tests?
- What are the ethical issues, and do I need approval?
- What are the limitations, and how will I acknowledge them?
- Have I justified every choice with a reference?
When in doubt, take a draft of this to your supervisor — early feedback prevents the costly mistake of collecting the wrong data.
Writing the Chapter Well
Keep the prose clear and the structure predictable: move from philosophy to design, then data collection, sampling, analysis, and ethics — the same outside-in logic as the Research Onion. Write in the past tense if the research is done, justify each decision as you introduce it, and support claims with scholarly sources. A clean, well-argued methodology chapter is often the section where careful students pull clearly ahead.
Frequently Asked Questions
How long should a master’s methodology chapter be?
Typically 2,000–3,000 words, or roughly 15–20% of a dissertation, though always check your own programme’s guidance — word counts and weighting vary by university and department.
Which is better, qualitative or quantitative?
Neither is inherently better; the right choice depends entirely on your research questions. Use quantitative methods to measure and test, qualitative to understand and explore, and mixed methods when you need both.
Do I have to use the Saunders Research Onion?
It isn’t compulsory, but it’s the framework most UK supervisors expect, and structuring your chapter around its layers makes your reasoning easy to follow and easy to mark. Check what your department prefers.
What’s the difference between methodology and methods?
Methods are the specific tools you use (interviews, surveys, regression). Methodology is the wider justification — the philosophy, approach, and reasoning that explain why those methods are appropriate.
Do I need ethical approval?
If your study involves human participants, almost always yes — and you must secure it before collecting data. Studies using only publicly available secondary data often need less, but confirm with your supervisor and ethics committee.
Can I use only secondary data?
Yes. A rigorous secondary-data study (using datasets, reports, or a systematic literature review) is a fully legitimate route, and a sensible one when time or access to participants is limited.
Choosing your methodology is the decision that shapes every chapter that follows it. Get the alignment right — questions, philosophy, methods, analysis — and the rest of your dissertation has a foundation it can stand on.
